A method and system for optimizing network fault traffic based on security protocol scheduling

By setting up test points in multiple geographical locations and utilizing security protocol mechanisms to redirect user traffic to a dynamic content delivery network (DCDN) during network failures and switch back to the public network when normal operations resume, this solves the problems of untimely network failure detection and high DCDN costs in existing technologies, achieving efficient and reliable network failure traffic optimization.

CN120602124BActive Publication Date: 2026-11-17ZUOYEBANG EDUCATION TECH (BEIJING) CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510654757.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-05-21
Publication Date
2026-11-17
Estimated Expiration
2045-05-21

AI Technical Summary

Technical Problem

Existing network fault detection methods are not timely or accurate enough, making it impossible to quickly switch to a backup network entry point when a network failure occurs, which affects business continuity and user experience. At the same time, dynamic content delivery network solutions are too costly.

Method used

By setting up test points in multiple geographical locations, the network in the target area can be monitored in real time. When the network fails, user traffic can be diverted to the dynamic content delivery network using security protocol mechanisms, and when the network recovers, traffic can be switched back to the public network through a load balancer, thus achieving flexible scheduling and reducing costs.

Benefits of technology

It enables timely detection of network faults and flexible traffic scheduling, reduces network operating costs, improves business continuity and user experience, reduces manual intervention, and ensures network efficiency and reliability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120602124B_ABST
    Figure CN120602124B_ABST
Patent Text Reader

Abstract

This invention provides a network fault traffic optimization method and system based on security protocol scheduling, belonging to the field of network communication technology. It includes: real-time monitoring of the network in a target area by setting up testing points in multiple geographical locations; when a network fault is detected in the target area, user traffic is scheduled to a dynamic content delivery network (DCN) based on a security protocol mechanism, and user devices access network links through the DCN; when the network in the target area is detected to have returned to normal, user traffic is evenly distributed to multiple servers or network links through a load balancer based on the security protocol mechanism. This technical solution reduces network operating costs by using the DCN only during network faults, and improves network operating efficiency and reliability by intelligently switching between the public network and the DCN based on a security protocol mechanism, ensuring business continuity and enhancing user experience.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of network communication technology, and in particular to a method and system for optimizing network fault traffic based on security protocol scheduling. Background Technology

[0002] As businesses expand globally, they frequently encounter network jitter issues, especially in certain regions where carrier failures can cause widespread network instability, leading to service interruptions and customer churn. To address this, businesses typically employ solutions from cloud providers and other vendors' dynamic content delivery networks (CDNs). These solutions improve network availability and reduce latency through dedicated lines and redundancy, but they are significantly more expensive, costing more than twice the price of normal public network bandwidth.

[0003] While existing Dynamic Content Delivery Network (DCDN) solutions offer excellent availability and latency, their high cost is a major technical drawback. Specifically, DCDN solutions ensure network stability through dedicated lines and line redundancy, but this approach is prohibitively expensive for many enterprises. Furthermore, existing network fault detection methods are often not timely or accurate enough, preventing rapid switching to backup network access points in the event of a failure. This impacts business continuity and negatively affects the user's network experience.

[0004] Therefore, a network fault traffic optimization method and system based on security protocol scheduling is proposed. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a network fault traffic optimization method and system based on security protocol scheduling, which solves the problems of poor performance in existing network fault detection and traffic scheduling, as well as the high cost of dynamic content delivery network solutions, thereby improving service stability and user experience.

[0006] This invention provides a method for optimizing network fault traffic based on security protocol scheduling, comprising:

[0007] By setting up test points in multiple geographical locations, the network in the target area can be monitored in real time;

[0008] When a network failure is detected in the target area, user traffic is redirected to a dynamic content delivery network based on a security protocol mechanism, and user devices access network links through the dynamic content delivery network.

[0009] When the network in the target area is detected to have returned to normal, the user traffic is evenly distributed to multiple servers or network links through a load balancer based on a security protocol mechanism, and the use of the dynamic content distribution network is disconnected.

[0010] Preferably, the present invention provides a network fault traffic optimization method based on security protocol scheduling, wherein the steps include: real-time monitoring of the network in the target area by setting up test points in multiple geographical locations; including:

[0011] Test points are deployed in multiple different geographical locations, and each test point is equipped with a corresponding test terminal.

[0012] The network monitoring terminal sends a testing task to the testing terminal to perform real-time network monitoring in the target area and sends the obtained testing results back to the network monitoring terminal; including:

[0013] Based on the deployment status of the testing terminals, the network monitoring terminal formulates a testing plan and sends the testing tasks to the testing terminals.

[0014] The testing terminal parses and processes the planned testing task to obtain the testing start time, testing cycle, testing end time, and number of polling. When the current time period reaches the testing start time, the testing terminal is woken up to monitor the network of the target area based on the planned testing task. When the current time period reaches the testing end time, the testing terminal sends the terminal monitoring results back to the network monitoring terminal and controls the testing terminal to enter a sleep state.

[0015] The network monitoring terminal parses and processes the returned terminal monitoring results. When no network abnormality is detected, the terminal monitoring results are used as the dial test results. When a network abnormality is detected, the network monitoring terminal sends an instant dial test task to the dial test terminals in the target area network to monitor the network simultaneously. The dial test results are obtained based on the instant test results returned by each dial test terminal.

[0016] Preferably, the present invention provides a network fault traffic optimization method based on security protocol scheduling, wherein the steps include: when a network fault is detected in the target area, user traffic is scheduled to a dynamic content delivery network based on a security protocol mechanism, and the user terminal device accesses the network link through the dynamic content delivery network; including:

[0017] When a network failure is detected in the target area, the request information and destination address are obtained by parsing user traffic. The request information and destination address are then encapsulated into an information request packet using an encryption protocol based on a security protocol mechanism.

[0018] The information request packet is transmitted to the protocol parser via the dynamic content delivery network.

[0019] The information request packet is subjected to security detection through the monitoring port on the protocol parser, and the request detection result is obtained.

[0020] After the security check of the information request packet is passed, the received information request packet is parsed to obtain the request information and the destination address, and the request information is transmitted to the target server according to the destination address.

[0021] Preferably, the present invention provides a network fault traffic optimization method based on security protocol scheduling, wherein the step of: transmitting the information request packet to the protocol parser through the dynamic content delivery network includes:

[0022] Based on the information request packet, obtain the network server cluster with corresponding processing capabilities in the dynamic content delivery network;

[0023] A search area is established based on the network server cluster. Several search points are set, and the search points randomly select initial positions in the search area. The spatial positions are updated according to the network server's performance evaluation value to obtain the sub-search area of ​​each search point.

[0024]

[0025] in, R is the update position for the i-th search point. best The current optimal position is denoted by 'a', where 'a' is the position change parameter and 'K' is the position. j R is the parameter for the sine mapping. m This represents the current average distribution location of all search points. This represents the current position of the i-th search point;

[0026]

[0027] Among them, c j K is a random parameter in the parameter interval [0, C]. j-1 The sine mapping parameters are the positions of the last updated positions;

[0028] The search point updates its position in the sub-search area using a first search path and obtains a first optimized network server.

[0029]

[0030] in, For the i-th search point in the corresponding sub-search region z i The update position obtained using the first search path, w i Here, represents the perturbation parameter value corresponding to the i-th search point, and rand is a random parameter. For the i-th search point in the corresponding sub-search region z i The optimal position obtained by the first search path, Δx is the lateral displacement of the search point based on the current position and the updated position, and Δy is the longitudinal displacement of the search point based on the current position and the updated position.

[0031] The search point updates its position in the sub-search area using the second search path and obtains the second optimized network server.

[0032]

[0033] in, For the i-th search point in the corresponding sub-search region z i The update position obtained using the second search path. For the i-th search point in the corresponding sub-search region z i The location of the network server corresponding to the maximum performance evaluation value obtained by searching using the second search path. For the i-th search point in the corresponding sub-search region z i The location of the network server corresponding to the minimum performance evaluation value obtained by searching using the second search path. For the i-th search point in the corresponding sub-search region z i The current position obtained using the second search path;

[0034] The optimal network server is obtained by comparing the first and second performance evaluation values ​​corresponding to the first and second optimized network servers.

[0035] When the optimal network server is obtained, the transmission distance between the optimal network server and the user terminal device is evaluated. When the transmission distance meets the threshold requirement, the information request packet is transmitted through the optimal network server.

[0036] Preferably, the present invention provides a network fault traffic optimization method based on security protocol scheduling, the steps of which include: establishing a search region based on the network server cluster, setting several search points, randomly selecting initial positions for the search points within the search region, and updating the spatial positions according to the network server's performance evaluation values ​​to obtain a sub-search region for each search point; including:

[0037] Based on the operating status of the network server and the network quality, obtain the working performance evaluation value of the network server;

[0038] S i =μ1*(α) d *d i +α c *ci +α m *m i )+μ2*(σ t *t i +σ p *p i )

[0039] Among them, S i Let μ1 be the performance evaluation value of the i-th network server, μ2 be the first weight parameter, and μ2 be the second weight parameter; d i Let α be the bandwidth remaining rate of the i-th network server. d c represents the corresponding bandwidth weight value. i Let α be the remaining computing resources of the i-th network server. c For the corresponding resource weight value, m i The remaining memory percentage of the i-th network server, α m For the corresponding memory weight value; t i The transmission delay for communicating with the i-th network server, σ t p represents the corresponding delay weight value. i Let σ be the packet loss rate of the i-th network server. p This represents the corresponding packet loss weight value.

[0040] Preferably, the present invention provides a network fault traffic optimization method based on security protocol scheduling, wherein the step of: performing security detection on information request packets through the monitoring port on the protocol parser and obtaining the request detection result; includes:

[0041] Obtain the transmission delay, transmission direction, and data length of the information request packet, and construct a detection sequence;

[0042] The detection sequence is preprocessed and shaped according to the model detection requirements. The dimensions of the detection sequence are adjusted, the continuous features of the information request packet are arranged in the horizontal direction, and the information requests of the information request packet at different times are arranged in the vertical direction to construct the preprocessed detection sequence.

[0043] Spatial feature information of the preprocessed detection sequence is extracted by a convolutional neural network model, temporal feature information of the preprocessed detection sequence is extracted by a long short-term memory network model, key feature information of the preprocessed detection sequence is extracted by a multi-attention mechanism network model, and feature mapping is performed through a fully connected layer to obtain detection feature information.

[0044] A security detection model is constructed, sample labels are added to the sample detection sequence, the security detection model is trained based on the loss function, and the control parameters of the security detection model are optimized.

[0045]

[0046] Where, θ i ′ θ is the control parameter obtained by optimization based on the detection sequence of the i-th sample. 0 Here, γ represents the initial control parameters of the security detection model, γ is the learning rate on the sample detection sequence, and L... i (θ 0 ) is based on the initial control parameter θ 0 Set the loss function value obtained by the security detection model for the detection sequence of the i-th sample;

[0047] The control parameters of the security detection model are optimized by using all sample detection sequences to obtain optimized control parameters;

[0048]

[0049] Where θ is the optimization control parameter of the security detection model, and there are a total of n sample detection sequences;

[0050] The detection feature information is input into the security detection model for detection, and the corresponding request detection results are obtained. When the detection sequence is detected as malicious access, blocking measures are implemented to prevent the protocol parser from parsing and processing the information request packet.

[0051] Preferably, the present invention provides a network fault traffic optimization method based on security protocol scheduling, wherein the steps include: when the network in the target area is detected to have returned to normal, the user traffic is evenly distributed to multiple servers or network links through a load balancer based on a security protocol mechanism, and the use of the dynamic content delivery network is disconnected; including:

[0052] The load balancer receives user traffic transmitted from user devices, parses it, and obtains request information and destination address.

[0053] The load balancer obtains public servers capable of processing the request information based on the request information and constructs the public server pool.

[0054] By obtaining the processor resources of the public network servers in the public network server pool, the response time of the public network server to the request information is calculated, and the information transmission time is calculated according to the transmission route between the load balancer and the public network server. Based on the response time and the information transmission time, the static processing response time is obtained.

[0055] Obtain the real-time task processing status of the public servers in the current public server pool, and obtain the dynamic processing response time of the public servers to the request information;

[0056] Based on the static processing response time and dynamic processing response time, the processing response time corresponding to each public network server is obtained. The load balancer selects the public network server with the shortest processing response time as the request processing server, modifies the destination address to the access address of the request processing server, and transmits the user traffic to the request processing server.

[0057] Preferably, the present invention provides a network fault traffic optimization method based on security protocol scheduling, comprising:

[0058] The load balancer provides a unique access point for user devices to receive user traffic.

[0059] The security protocol mechanism caches the access points of the load balancer and the dynamic content distribution network, and automatically switches the access points of the load balancer and the dynamic content distribution network according to the network operation status of the target area.

[0060] The user terminal device caches the access entry point of the dynamic content delivery network. When the network request fails, the user terminal device automatically retryes the network access through the access entry point of the dynamic content delivery network.

[0061] This invention provides a network fault traffic optimization system based on security protocol scheduling, comprising:

[0062] The network monitoring module is used to monitor the network in a target area in real time by setting up test points in multiple geographical locations;

[0063] The network scheduling module is used to schedule user traffic to the dynamic content delivery network based on a security protocol mechanism when a network failure is detected in the target area. The user terminal device accesses the network link through the dynamic content delivery network.

[0064] The network scheduling module is also used to, when the network in the target area is detected to have returned to normal, distribute user traffic evenly to multiple servers or network links through a load balancer based on a security protocol mechanism, and disconnect the use of the dynamic content distribution network.

[0065] Compared with traditional technologies, the advantages of this invention are as follows: A network fault traffic optimization method and system based on security protocol scheduling achieves real-time automatic fault monitoring of the target area network by setting up testing points in multiple geographical locations. When a network fault is detected, user traffic is scheduled to a dynamic content delivery network (DCN) based on a security protocol mechanism to ensure the continuity of user services. When the network returns to normal, it switches back to the normal public network, achieving flexible scheduling of user traffic and avoiding unnecessary network costs. The above technical solution reduces network operating costs by using the DCN only during network faults, avoiding the problem of high costs associated with long-term use of the DCN. At the same time, the intelligent switching between the public network and the DCN based on the security protocol mechanism reduces manual intervention, improves network operating efficiency and reliability, ensures service continuity, and enhances user experience.

[0066] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0067] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0068] Figure 1 This is a flowchart illustrating the network fault traffic optimization method based on security protocol scheduling provided by the present invention.

[0069] Figure 2 This is a schematic diagram of the network fault traffic optimization system based on security protocol scheduling provided by the present invention. Detailed Implementation

[0070] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0071] Example 1:

[0072] This invention provides a method for optimizing network fault traffic based on security protocol scheduling, referencing... Figure 1 ,include:

[0073] By setting up test points in multiple geographical locations, the network in the target area can be monitored in real time;

[0074] When a network failure is detected in the target area, user traffic is redirected to the dynamic content delivery network based on a security protocol mechanism, and user devices access the network link through the dynamic content delivery network.

[0075] When the network in the target area is detected to have returned to normal, the user traffic is evenly distributed to multiple servers or network links through a load balancer based on the security protocol mechanism, and the use of the dynamic content delivery network is disconnected.

[0076] In the above embodiments, by setting up test points in multiple geographical locations, the network in the target area is monitored in real time. When a network failure is detected in the target area, user traffic is redirected to the dynamic content delivery network based on a security protocol mechanism, and the user terminal device accesses the network link through the dynamic content delivery network. When the network in the target area is detected to have returned to normal, the user traffic is evenly distributed to multiple servers or network links through a load balancer based on a security protocol mechanism, and the use of the dynamic content delivery network is disconnected.

[0077] In the above embodiments, the security protocol mechanism is implemented as the DOH (DNS-over-HTTPS) protocol. The DNS query request is transmitted using the HTTPS protocol. By encapsulating the DNS query in HTTPS, the problem of user DNS query requests being eavesdropped on or modified in the original DNS protocol is avoided, thereby enhancing the security and privacy of DNS queries.

[0078] In the above embodiments, the Dynamic Content Delivery Network (DCDN) is a network used to accelerate the transmission of dynamic content. It not only caches static content but also reduces the distance between the user and the origin server by optimizing dynamic request paths, intelligent scheduling, and edge computing, thereby improving the transmission efficiency of dynamic content. It is primarily used to accelerate dynamic content, such as database query results, API request responses, and personalized pages generated in real time.

[0079] In the above embodiments, a load balancer (LB) is a device or software used to evenly distribute network traffic or computing tasks across multiple servers or network links. It provides a single entry point to the outside world through a proxy server, and forwards requests to an internal server cluster based on request characteristics and a preset algorithm, thereby improving system concurrency, availability, and response speed.

[0080] The beneficial effects of the above technologies are as follows: By setting up testing points in multiple geographical locations, real-time automatic fault monitoring of the target area network is achieved. When a network failure is detected, user traffic is diverted to the Dynamic Content Delivery Network (DCN) based on a security protocol mechanism to ensure the continuity of user services. When the network returns to normal, it switches back to the normal public network, achieving flexible scheduling of user traffic and avoiding unnecessary network costs. Compared with traditional technologies, the above methods reduce network operating costs by using the DCN only during network failures, avoiding the problem of high costs associated with long-term use of the DCN. At the same time, the intelligent switching between the public network and the DCN based on the security protocol mechanism reduces manual intervention, improves network operating efficiency and reliability, ensures business continuity, and enhances user experience.

[0081] Example 2:

[0082] This invention provides a method for optimizing network fault traffic based on security protocol scheduling, comprising the following steps: real-time monitoring of the network in the target area by setting up testing points in multiple geographical locations; including:

[0083] Test points are deployed in multiple different geographical locations, and each test point is equipped with a corresponding test terminal.

[0084] The network monitoring terminal sends a dial-up test task to the dial-up test terminal to perform real-time network monitoring in the target area and transmits the obtained test results back to the network monitoring terminal; including:

[0085] Based on the deployment of the testing terminals, the network monitoring terminal formulates a testing plan and sends the testing tasks to the testing terminals.

[0086] The testing terminal parses and processes the planned testing tasks to obtain the testing start time, testing cycle, testing end time and number of polling. When the testing start time is reached in the current time period, the testing terminal is woken up to monitor the network of the target area based on the planned testing tasks. When the testing end time is reached in the current time period, the testing terminal sends the terminal monitoring results back to the network monitoring terminal and controls the testing terminal to enter a sleep state.

[0087] The network monitoring terminal parses and processes the returned terminal monitoring results. When no network anomalies are detected, the terminal monitoring results are used as the dial test results. When network anomalies are detected, the network monitoring terminal sends an instant dial test task to the dial test terminals in the target area network to monitor the network in the target area simultaneously. The dial test results are obtained based on the instant test results returned by each dial test terminal.

[0088] In the above embodiments, test points are deployed in multiple different geographical locations, and each test point is equipped with a corresponding test terminal. Test tasks are sent from the network monitoring terminal to the test terminal to perform real-time network detection in the target area, and the obtained test results are sent back to the network monitoring terminal.

[0089] In the above embodiments, the network monitoring terminal sends the planned testing task to the testing terminal. The testing terminal parses and processes the planned testing task to obtain the testing start time, testing cycle, testing end time and polling count. When the current time period reaches the testing start time, the testing terminal is woken up to monitor the network of the target area based on the planned testing task. When the current time period reaches the testing end time, the testing terminal sends the terminal monitoring results back to the network monitoring terminal and controls the testing terminal to enter a sleep state.

[0090] In the above embodiments, the network monitoring terminal parses and processes the returned terminal monitoring results. When an anomaly is detected in the network, the network monitoring terminal sends an instant dial-up test task to the dial-up test terminal of the network in the target area to monitor the network simultaneously. Based on the instant test results returned by each dial-up test terminal, the dial-up test results are obtained.

[0091] The beneficial effects of the above technologies are as follows: By issuing planned testing tasks to the testing terminals through the network monitoring terminal, the network testing terminals can perform polling monitoring of the target area network, avoiding the problem of multiple testing terminals continuously monitoring the network for a long time and consuming resources. Furthermore, when any testing terminal detects a network anomaly, all testing terminals are awakened to perform real-time network monitoring and obtain the testing results. The above technical solution can detect network faults in a timely and accurate manner, providing a basis for subsequent traffic scheduling. While realizing the monitoring of the target area network, it effectively reduces the power consumption required for monitoring.

[0092] Example 3:

[0093] This invention provides a network fault traffic optimization method based on security protocol scheduling, comprising the following steps: when a network fault is detected in a target area, user traffic is scheduled to a dynamic content delivery network based on a security protocol mechanism, and user devices access network links through the dynamic content delivery network; including:

[0094] When a network failure is detected in the target area, the request information and destination address are obtained by parsing user traffic. The request information and destination address are then encapsulated into an information request packet using an encrypted protocol based on a security protocol mechanism.

[0095] The message request packet is transmitted to the protocol parser via a dynamic content delivery network;

[0096] The security check of the information request packet is performed through the monitoring port on the protocol resolver, and the request check result is obtained.

[0097] After the security check of the information request packet is passed, the received information request packet is parsed to obtain the request information and destination address, and the request information is transmitted to the target server according to the destination address.

[0098] In the above embodiments, when a network failure is detected in the target area, the request information and destination address are obtained based on user traffic parsing. The request information and destination address are encapsulated into an information request packet using an encryption protocol based on a security protocol mechanism. The information request packet is transmitted to the protocol parser through a dynamic content delivery network. After the monitoring port on the protocol parser passes the security test of the information request packet, it parses the received information request packet to obtain the request information and destination address. The request information is then transmitted to the target server based on the destination address.

[0099] In the above embodiments, the security detection of information request packets is achieved through the monitoring port on the protocol parser, and the parsing process of information request packets is stopped when the security detection fails.

[0100] The beneficial effects of the above technologies are as follows: Dynamic content delivery networks enable the scheduling and processing of user traffic when network failures occur; user traffic is parsed and encapsulated into information request packets using encrypted protocols based on security mechanisms and transmitted to the protocol parser; the request information is then transmitted to the target server according to the destination address, thus realizing the data transmission function of dynamic content delivery networks. Furthermore, monitoring ports are set up to detect information request packets, avoiding the processing of information request packets containing virus files, effectively improving the security of information transmission.

[0101] Example 4:

[0102] This invention provides a method for optimizing network fault traffic based on security protocol scheduling, comprising the steps of: transmitting information request packets to a protocol parser via a dynamic content delivery network; including:

[0103] Based on the information request packet, obtain the network server cluster with corresponding processing capabilities in the dynamic content delivery network;

[0104] A search region is established based on a network server cluster. Several search points are set, and the search points randomly select initial positions within the search region. The spatial positions are updated based on the network server's performance evaluation values ​​to obtain the sub-search region for each search point.

[0105]

[0106] in, R is the update position for the i-th search point. best The current optimal position is denoted by 'a', where 'a' is the position change parameter and 'K' is the position.j R is the parameter for the sine mapping. m This represents the current average distribution location of all search points. This represents the current position of the i-th search point;

[0107]

[0108] Among them, c j K is a random parameter in the parameter interval [0, C]. j-1 The sine mapping parameters are the positions of the last updated positions;

[0109] The search point updates its position in the sub-search area using the first search path and obtains the first optimized network server.

[0110]

[0111] in, For the i-th search point in the corresponding sub-search region z i The update position obtained using the first search path, w i Here, represents the perturbation parameter value corresponding to the i-th search point, and rand is a random parameter. For the i-th search point in the corresponding sub-search region z i The optimal position obtained by the first search path, Δx is the lateral displacement of the search point based on the current position and the updated position, and Δy is the longitudinal displacement of the search point based on the current position and the updated position.

[0112] The search point updates its position in the sub-search area using the second search path and obtains the second optimized network server.

[0113]

[0114] in, For the i-th search point in the corresponding sub-search region z i The update position obtained using the second search path. For the i-th search point in the corresponding sub-search region z i The location of the network server corresponding to the maximum performance evaluation value obtained by searching using the second search path. For the i-th search point in the corresponding sub-search region z i The location of the network server corresponding to the minimum performance evaluation value obtained by searching using the second search path. For the i-th search point in the corresponding sub-search region z i The current position obtained using the second search path;

[0115] The optimal network server is obtained by comparing the first and second performance evaluation values ​​corresponding to the first and second optimized network servers.

[0116] When the optimal network server is obtained, the transmission distance between the optimal network server and the user terminal device is evaluated. If the transmission distance meets the threshold requirement, the information request packet is transmitted through the optimal network server.

[0117] In the above embodiments, the dynamic content delivery network obtains a cluster of network servers with corresponding processing capabilities in the dynamic content delivery network based on the information request packet, selects the optimal network server in the network server cluster, and when the transmission distance between the optimal network server and the user terminal device meets the threshold requirement, the information request packet is transmitted to the target server through the optimal network server.

[0118] In the above embodiments, for the selection of the optimal network server, a search area is established based on the network server cluster. By setting several search points, the search points randomly select initial positions in the search area and update their spatial positions according to the network server's performance evaluation value to obtain a sub-search area for each search point. The search points update their positions in the sub-search areas using the first search path and the second search path respectively to obtain the first optimized network server and the second optimized network server. The optimal network server is then obtained based on the corresponding performance evaluation.

[0119] In the above embodiments, the sub-search region of the search point is obtained by adding a sine mapping parameter to the search point based on its current position and the currently obtained optimal position. The setting of the sine mapping parameter effectively improves the global search capability of the search point.

[0120] In the above embodiments, the search point updates its position in the sub-search region using the first search path, and by adding random perturbation, the convergence speed of the algorithm is accelerated, while the risk of getting trapped in local optima during the search process is reduced, and the first optimized network server of the sub-search region is obtained.

[0121] In the above embodiments, the search point updates its position in the sub-search region using the second search path. By introducing a reverse learning strategy to seek the optimal solution in the sub-search region, the optimization accuracy of the algorithm is effectively improved.

[0122] The beneficial effects of the above technologies are as follows: by acquiring a network server cluster capable of processing information request packets, setting a search point to search for the optimal network server in the network server cluster, and transmitting the information request packet to the target server through the optimal network server when the transmission distance meets the threshold requirement, the optimal network server is intelligently selected for the transmission of information request packets in the dynamic content delivery network. Furthermore, by setting the threshold requirement for transmission distance, the processing rate of user traffic in the dynamic content delivery network is further improved.

[0123] Example 5:

[0124] This invention provides a method for optimizing network fault traffic based on security protocol scheduling. The steps include: establishing a search region based on a network server cluster; setting several search points; randomly selecting initial positions for each search point within the search region; updating the spatial positions based on the network server's performance evaluation values; and obtaining a sub-search region for each search point.

[0125] Based on the network server's operating status and network quality, obtain the network server's performance evaluation value;

[0126] S i =μ1*(α) d *d i +α c *c i +α m *m i )+μ2*(τ t *t i +τ p *p i )

[0127] Among them, S i Let μ1 be the performance evaluation value of the i-th network server, μ2 be the first weight parameter, and μ2 be the second weight parameter; d i Let α be the bandwidth remaining rate of the i-th network server. d c represents the corresponding bandwidth weight value. i Let α be the remaining computing resources of the i-th network server. c For the corresponding resource weight value, m i The remaining memory percentage of the i-th network server, α m For the corresponding memory weight value; t i The transmission delay for communicating with the i-th network server, σ t p represents the corresponding delay weight value. i Let σ be the packet loss rate of the i-th network server. p This represents the corresponding packet loss weight value.

[0128] In the above embodiments, the local working performance of the network server is calculated by obtaining the bandwidth remaining rate, computing resource remaining rate, and memory remaining rate of the network server. The transmission status performance of the network server is calculated by obtaining the transmission latency and packet loss rate of the network server during communication. By dynamically allocating weights, the evaluation value of the working performance of the network server is obtained.

[0129] The beneficial effects of the above technologies are as follows: by analyzing the operating status and network quality of network servers, the performance evaluation value of network servers can be obtained, thereby providing a reliable basis for searching for the optimal network server in the network server cluster.

[0130] Example 6:

[0131] This invention provides a method for optimizing network fault traffic based on security protocol scheduling, comprising the following steps: performing security checks on information request packets through a monitoring port on a protocol resolver, and obtaining the request detection results; including:

[0132] Obtain the transmission delay, transmission direction, and data length of the information request packet, and construct a detection sequence;

[0133] According to the model detection requirements, the detection sequence is preprocessed and shaped, the dimensions of the detection sequence are adjusted, the continuous features of the information request packet are arranged in the horizontal direction, and the information requests of the information request packet at different times are arranged in the vertical direction to construct the preprocessed detection sequence.

[0134] Spatial feature information of the preprocessed detection sequence is extracted by a convolutional neural network model, temporal feature information of the preprocessed detection sequence is extracted by a long short-term memory network model, key feature information of the preprocessed detection sequence is extracted by a multi-attention mechanism network model, and feature mapping is performed through a fully connected layer to obtain detection feature information.

[0135] A security detection model is constructed by adding sample labels to the sample detection sequences, training the security detection model based on the loss function, and optimizing the control parameters of the security detection model.

[0136]

[0137] Where, θ i ′ θ is the control parameter obtained by optimization based on the detection sequence of the i-th sample. 0 Here, γ represents the initial control parameters of the security detection model, γ is the learning rate on the sample detection sequence, and L is the learning rate. i (θ 0 ) is based on the initial control parameter θ 0 Set the loss function value obtained by the security detection model for the detection sequence of the i-th sample;

[0138] The control parameters of the security detection model are optimized by using all sample detection sequences to obtain the optimized control parameters;

[0139]

[0140] Where θ is the optimization control parameter of the security detection model, and there are a total of n sample detection sequences;

[0141] The detection feature information is input into the security detection model for detection, and the corresponding request detection results are obtained. When the detection sequence is detected as malicious access, blocking measures are implemented to prevent the protocol parser from parsing and processing the information request packet.

[0142] In the above embodiments, the transmission delay, transmission direction, and data length of the information request packet are obtained to construct a detection sequence; the detection sequence is preprocessed and shaped according to the model detection requirements to construct a preprocessed detection sequence; spatial feature information of the preprocessed detection sequence is extracted through a convolutional neural network model, temporal feature information of the preprocessed detection sequence is extracted through a long short-term memory network model, and key feature information of the preprocessed detection sequence is extracted through a multi-attention mechanism network model; feature mapping is performed through a fully connected layer to obtain detection feature information; the detection feature information is input into the security detection model for detection to obtain the corresponding request detection result; when the detection sequence is detected as malicious access, blocking measures are implemented to prohibit the protocol parser from parsing and processing the information request packet.

[0143] In the above embodiments, the preprocessing and shaping of the detection sequence is achieved by adjusting the dimensions of the detection sequence. Continuous features of the information request packet are arranged in the horizontal direction, and information requests at different times are arranged in the vertical direction to construct the preprocessed detection sequence. The content of the preprocessed detection sequence data remains unchanged, only the shape of the data dimension is changed, which facilitates the movement of the convolution kernel in the model in subsequent steps to capture more comprehensive data feature relationships.

[0144] In the above embodiments, sample labels are added to the sample detection sequence, the security detection model is trained based on the loss function, the control parameters of the security detection model are optimized, and the optimized security detection model is used for security detection of the preprocessed detection sequence.

[0145] In the above embodiments, the loss function is constructed based on the cross-entropy loss function. It is trained by multiple sample detection sequences so that the control parameters of the security detection model obtain the most universal optimization direction, thereby improving the generalization performance of the model on security detection tasks.

[0146] The beneficial effects of the above technologies are as follows: They acquire the transmission delay, transmission direction, and data length of information request packets, construct detection sequences, perform preprocessing and shaping, obtain preprocessed detection sequences, and use these sequences to extract detection feature information through convolutional neural network models, long short-term memory network models, and multi-attention mechanism network models. Furthermore, they perform security detection through security detection models, obtain request detection results, and achieve secure detection of information request packets by the monitoring port. This prevents dynamic content delivery networks from forwarding and parsing maliciously accessed data, thus avoiding impacts on the normal operation of the server. While achieving secure detection of user traffic, it also improves the security protection of the server.

[0147] Example 7:

[0148] This invention provides a network fault traffic optimization method based on security protocol scheduling, comprising the following steps: when the network in the target area is detected to have returned to normal, based on a security protocol mechanism, user traffic is evenly distributed to multiple servers or network links through a load balancer, and the use of the dynamic content delivery network is disconnected; including:

[0149] The load balancer receives user traffic transmitted from user devices, parses it, and obtains request information and destination address.

[0150] The load balancer obtains public servers capable of processing request information based on the request information and builds a public server pool.

[0151] By obtaining the processor resources of the public servers in the public server pool, the response time of the public servers to the request information is calculated, and the information transmission time is calculated according to the transmission route between the load balancer and the public servers. Based on the response time and information transmission time, the static processing response time is obtained.

[0152] Get the real-time task processing status of public servers in the current public server pool, and get the dynamic processing response time of public servers to request information;

[0153] Based on the static and dynamic processing response times, the processing response times of each public network server are obtained. The load balancer selects the public network server with the shortest processing response time as the request processing server, modifies the destination address to the access address of the request processing server, and transmits user traffic to the request processing server.

[0154] In the above embodiments, the load balancer receives user traffic transmitted from user devices, parses it, and obtains request information and destination address. Based on the request information, the load balancer obtains public network servers capable of processing request information, selects the public network server with the shortest processing response time as the request processing server, modifies the destination address to the access address of the request processing server, and transmits the user traffic to the request processing server.

[0155] In the above embodiments, the processing response time of the public network server is obtained by calculating the static processing response time and the dynamic processing response time. The static processing response time is calculated based on the public network server's response time to the request information and the information transmission time required for the transmission route between the load balancer and the public network server. The dynamic processing response time is obtained based on the current real-time task processing status of the public network server.

[0156] The beneficial effects of the above technologies are as follows: when the network in the target area is detected to have returned to normal, the load balancer obtains the response time of each public server in the public server pool to the processing of request information, selects the public server with the shortest processing response time as the request processing server to process the request information, and modifies the destination address to the access address of the request processing server, thereby improving the processing speed of request information, realizing the rapid scheduling and processing of request information, and effectively improving the network access experience of the user.

[0157] Example 8:

[0158] This invention provides a method for optimizing network fault traffic based on security protocol scheduling, comprising:

[0159] A load balancer provides a single access point for user devices to receive user traffic;

[0160] The security protocol mechanism caches the access points of the load balancer and the dynamic content delivery network, and automatically switches the access points of the load balancer and the dynamic content delivery network according to the network operation status of the target area.

[0161] The client device caches the access point of the dynamic content delivery network. When the client device fails to access the network, it automatically retryes the network access through the access point of the dynamic content delivery network.

[0162] In the above embodiments, when the network in the target area is fault-free, the user terminal device connects to the load balancer through a single access point, and the load balancer schedules user traffic.

[0163] In the above embodiments, the security protocol mechanism automatically switches the access points of the load balancer and the dynamic content delivery network according to the network failure situation, thereby realizing intelligent scheduling of user traffic.

[0164] In the above embodiments, by caching the access point of the dynamic content delivery network on the user terminal device, when a network request fails to be made through the public network, the network access will be automatically retried through the access point of the dynamic content delivery network.

[0165] The beneficial effects of the above technologies are as follows: By using a load balancer, user traffic can be scheduled when the target area network is fault-free; by using a dynamic content delivery network, user traffic can be processed when the target area network fails; and by intelligently switching between the two network scheduling methods based on a security protocol mechanism, normal processing of user traffic is effectively guaranteed, avoiding the impact of network failures on user network access; furthermore, the access entry point of the dynamic content delivery network is cached on the user-end device, and automatic switching occurs when network requests fail, effectively improving the user network access experience.

[0166] Example 9:

[0167] This invention provides a network fault traffic optimization system based on security protocol scheduling, referencing... Figure 2 ,include:

[0168] The network monitoring module is used to monitor the network in a target area in real time by setting up test points in multiple geographical locations;

[0169] The network scheduling module is used to schedule user traffic to the dynamic content delivery network based on a security protocol mechanism when a network failure is detected in the target area. User devices access network links through the dynamic content delivery network.

[0170] The network scheduling module is also used to distribute user traffic evenly to multiple servers or network links through a load balancer based on a security protocol mechanism when the network in the target area is detected to have returned to normal, thereby disconnecting the use of the dynamic content delivery network.

[0171] In the above embodiments, the network monitoring module monitors the network in the target area in real time by setting up test points in multiple geographical locations. When the network scheduling module detects a network failure in the target area, it schedules user traffic to the dynamic content delivery network based on a security protocol mechanism, and the user terminal device accesses the network link through the dynamic content delivery network. When the network in the target area is detected to have returned to normal, it distributes user traffic evenly to multiple servers or network links through a load balancer based on a security protocol mechanism, and disconnects the use of the dynamic content delivery network.

[0172] The beneficial effects of the above technologies are as follows: By setting up testing points in multiple geographical locations, real-time automatic fault monitoring of the target area network is achieved. When a network failure is detected, user traffic is diverted to the Dynamic Content Delivery Network (DCN) based on a security protocol mechanism to ensure the continuity of user services. When the network returns to normal, it switches back to the normal public network, achieving flexible scheduling of user traffic and avoiding unnecessary network costs. Compared with traditional technologies, the above system reduces network operating costs by using the DCN only during network failures, avoiding the problem of high costs associated with long-term use of the DCN. At the same time, the intelligent switching between the public network and the DCN based on the security protocol mechanism reduces manual intervention, improves network operating efficiency and reliability, ensures business continuity, and enhances user experience.

[0173] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A network fault traffic optimization method based on security protocol scheduling, characterized in that, include: By setting up test points in multiple geographical locations, the network in the target area can be monitored in real time; When a network failure is detected in the target area, user traffic is redirected to a dynamic content delivery network based on a security protocol mechanism, and user devices access network links through the dynamic content delivery network. When the network in the target area is detected to have returned to normal, the user traffic is evenly distributed to multiple servers or network links through a load balancer based on the security protocol mechanism, and the use of the dynamic content distribution network is disconnected. The load balancer provides a unique access point for user devices to receive user traffic. The security protocol mechanism caches the access points of the load balancer and the dynamic content distribution network, and automatically switches the access points of the load balancer and the dynamic content distribution network according to the network operation status of the target area. The user terminal device caches the access entry point of the dynamic content delivery network. When the network request fails, the user terminal device automatically retryes the network access through the access entry point of the dynamic content delivery network.

2. The network fault traffic optimization method based on security protocol scheduling according to claim 1, characterized in that, The steps include: setting up test points at multiple geographical locations to monitor the network in the target area in real time; including: Test points are deployed in multiple different geographical locations, and each test point is equipped with a corresponding test terminal. The network monitoring terminal sends a testing task to the testing terminal to perform real-time network monitoring in the target area and sends the obtained testing results back to the network monitoring terminal; including: Based on the deployment status of the testing terminals, the network monitoring terminal formulates a testing plan and sends the testing tasks to the testing terminals. The testing terminal parses and processes the planned testing task to obtain the testing start time, testing cycle, testing end time, and number of polling. When the current time period reaches the testing start time, the testing terminal is woken up to monitor the network of the target area based on the planned testing task. When the current time period reaches the testing end time, the testing terminal sends the terminal monitoring results back to the network monitoring terminal and controls the testing terminal to enter a sleep state. The network monitoring terminal parses and processes the returned terminal monitoring results. When no network abnormality is detected, the terminal monitoring results are used as the dial test results. When a network abnormality is detected, the network monitoring terminal sends an instant dial test task to the dial test terminals in the target area network to monitor the network simultaneously. The dial test results are obtained based on the instant test results returned by each dial test terminal.

3. The network fault traffic optimization method based on security protocol scheduling according to claim 1, characterized in that, The steps include: when a network failure is detected in the target area, user traffic is redirected to a dynamic content delivery network based on a security protocol mechanism, and user devices access network links through the dynamic content delivery network; including: When a network failure is detected in the target area, the request information and destination address are obtained by parsing user traffic. The request information and destination address are then encapsulated into an information request packet using an encryption protocol based on a security protocol mechanism. The information request packet is transmitted to the protocol parser via the dynamic content delivery network. The information request packet is subjected to security detection through the monitoring port on the protocol parser, and the request detection result is obtained. After the security check of the information request packet is passed, the received information request packet is parsed to obtain the request information and the destination address, and the request information is transmitted to the target server according to the destination address.

4. The network fault traffic optimization method based on security protocol scheduling according to claim 3, characterized in that, The step of transmitting the information request packet to the protocol parser via the dynamic content delivery network includes: Based on the information request packet, obtain the network server cluster with corresponding processing capabilities in the dynamic content delivery network; A search area is established based on the network server cluster. Several search points are set, and the search points randomly select initial positions in the search area. The spatial positions are updated according to the network server's performance evaluation value to obtain the sub-search area of ​​each search point. ; in, For the first Update location of each search point This is the optimal position obtained so far. For position change parameters, For sine mapping parameters, This represents the current average distribution location of all search points. For the first The current position of each search point; ; in, For parameter range random parameters, The sine mapping parameters are the positions of the last updated positions; The search point updates its position in the sub-search area using a first search path and obtains a first optimized network server. ; in, For the first Each search point is in its corresponding sub-search area. The update location obtained using the first search path. For the first The perturbation parameter values ​​corresponding to each search point For random parameters, For the first Each search point is in its corresponding sub-search area. The optimal position obtained using the first search path. The search point is the lateral displacement based on its current and updated positions. The vertical displacement of the search point based on the current position and the updated position; The search point updates its position in the sub-search area using the second search path and obtains the second optimized network server. ; in, For the first Each search point is in its corresponding sub-search area. The update position obtained using the second search path. For the first Each search point is in its corresponding sub-search area. The location of the network server corresponding to the maximum performance evaluation value obtained by searching using the second search path. For the first Each search point is in its corresponding sub-search area. The location of the network server corresponding to the minimum performance evaluation value obtained by searching using the second search path. For the first Each search point is in its corresponding sub-search area. The current position obtained using the second search path; The optimal network server is obtained by comparing the first and second performance evaluation values ​​corresponding to the first and second optimized network servers. When the optimal network server is obtained, the transmission distance between the optimal network server and the user terminal device is evaluated. When the transmission distance meets the threshold requirement, the information request packet is transmitted through the optimal network server.

5. The network fault traffic optimization method based on security protocol scheduling according to claim 4, characterized in that, The steps include: establishing a search region based on the network server cluster; setting several search points; randomly selecting initial positions for each search point within the search region; updating the spatial position based on the network server's performance evaluation value; and obtaining a sub-search region for each search point; including: Based on the operating status of the network server and the network quality, obtain the working performance evaluation value of the network server; ; in, For the first Performance evaluation values ​​for each network server As the first weight parameter, This is the second weighting parameter; For the first The bandwidth availability of each network server. For the corresponding bandwidth weight value, For the first The remaining computing resources of each network server. For the corresponding resource weight value, No. The remaining memory percentage of each network server. For the corresponding memory weight value; In order to be with the first The transmission latency for communication between network servers. For the corresponding delay weight value, For the first Packet loss rate of a network server This represents the corresponding packet loss weight value.

6. The network fault traffic optimization method based on security protocol scheduling according to claim 3, characterized in that, The steps include: performing security checks on the information request packets through the monitoring port on the protocol parser and obtaining the request detection results; including: Obtain the transmission delay, transmission direction, and data length of the information request packet, and construct a detection sequence; The detection sequence is preprocessed and shaped according to the model detection requirements. The dimensions of the detection sequence are adjusted, the continuous features of the information request packet are arranged in the horizontal direction, and the information requests of the information request packet at different times are arranged in the vertical direction to construct the preprocessed detection sequence. Spatial feature information of the preprocessed detection sequence is extracted by a convolutional neural network model, temporal feature information of the preprocessed detection sequence is extracted by a long short-term memory network model, key feature information of the preprocessed detection sequence is extracted by a multi-attention mechanism network model, and feature mapping is performed through a fully connected layer to obtain detection feature information. A security detection model is constructed, sample labels are added to the sample detection sequence, the security detection model is trained based on the loss function, and the control parameters of the security detection model are optimized. ; in, According to the first Control parameters obtained from optimizing individual sample detection sequences These are the initial control parameters for the security detection model. The learning rate on the sample detection sequence, To be based on initial control parameters Set up a security detection model for the first The loss function value is calculated for each sample detection sequence; The control parameters of the security detection model are optimized by using all sample detection sequences to obtain optimized control parameters; ; in, For the optimized control parameters of the safety detection model, there are a total of One sample detection sequence; The detection feature information is input into the security detection model for detection, and the corresponding request detection results are obtained. When the detection sequence is detected as malicious access, blocking measures are implemented to prevent the protocol parser from parsing and processing the information request packet.

7. The network fault traffic optimization method based on security protocol scheduling according to claim 1, characterized in that, The steps include: when the network in the target area is detected to have returned to normal, based on a security protocol mechanism, the user traffic is evenly distributed to multiple servers or network links through a load balancer, and the use of the dynamic content delivery network is disconnected; including: The load balancer receives user traffic transmitted from user devices, parses it, and obtains request information and destination address. The load balancer obtains public network servers capable of processing the request information based on the request information and constructs a public network server pool. By obtaining the processor resources of the public servers in the public server pool, the response time of the public server to the request information is calculated, and the information transmission time is calculated according to the transmission route between the load balancer and the public server. Based on the response time and the information transmission time, the static processing response time is obtained. Obtain the real-time task processing status of the public servers in the current public server pool, and obtain the dynamic processing response time of the public servers to the request information; Based on the static processing response time and dynamic processing response time, the processing response time corresponding to each public network server is obtained. The load balancer selects the public network server with the shortest processing response time as the request processing server, modifies the destination address to the access address of the request processing server, and transmits the user traffic to the request processing server.

8. A network fault traffic optimization system based on security protocol scheduling, characterized in that, include: The network monitoring module is used to monitor the network in a target area in real time by setting up test points in multiple geographical locations; The network scheduling module is used to schedule user traffic to the dynamic content delivery network based on a security protocol mechanism when a network failure is detected in the target area. The user terminal device accesses the network link through the dynamic content delivery network. The network scheduling module is also used to, when the network in the target area is detected to have returned to normal, distribute user traffic evenly to multiple servers or network links through a load balancer based on a security protocol mechanism, and disconnect the use of the dynamic content distribution network. The load balancer provides a unique access point for user devices to receive user traffic. The security protocol mechanism caches the access points of the load balancer and the dynamic content distribution network, and automatically switches the access points of the load balancer and the dynamic content distribution network according to the network operation status of the target area. The user terminal device caches the access entry point of the dynamic content delivery network. When the network request fails, the user terminal device automatically retryes the network access through the access entry point of the dynamic content delivery network.

Citation Information

Patent Citations

  • Area data transmission method, self-circulation domain name analysis system and self-circulation domain name analysis method

    CN108449444A

  • Method and system for accelerated scheduling of cloud disk service

    CN115883657A