A method for parsing abnormal DNS requests

By establishing exception standard indicators and game theory models, analyzing DNS request characteristics and predicting future requests, the problem of poor resolution of abnormal DNS requests in the existing technology is solved, and the adaptability and accuracy of network security protection strategies are improved.

CN117221273BActive Publication Date: 2024-11-08HUANENG INFORMATION TECH CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310980428.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-08-04
Publication Date
2024-11-08
Estimated Expiration
2043-08-04

AI Technical Summary

Technical Problem

In the prior art, the resolution accuracy of abnormal DNS requests is poor, resulting in low adaptability of network security protection strategies and often misjudgment and misjudgment problems.

Method used

By obtaining the logs of previous abnormal DNS requests, establishing multiple exception standard indicators, analyzing the characteristic data of DNS requests to be detected, using the Markov model to predict future requests, and establishing current and future game theory models to adjust network protection strategies to improve parsing accuracy and strategy adaptability.

Benefits of technology

It improves the resolution accuracy of abnormal DNS requests, ensures the adaptability and effectiveness of network protection policies, and reduces misjudgment and misjudgment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117221273B_ABST
    Figure CN117221273B_ABST
Patent Text Reader

Abstract

The present invention discloses an abnormal DNS request parsing method, which relates to the technical field of data parsing, including obtaining previous abnormal DNS requests, and establishing multiple abnormal standard indicators according to the logs of previous abnormal DNS requests; obtaining a DNS request to be detected, and generating multiple abnormal indicators according to the logs of the DNS request to be detected; comparing the abnormal indicators and the abnormal standard indicators, so as to determine whether the DNS request to be detected is an abnormal DNS request; if the DNS request to be detected is an abnormal DNS request, analyzing the request message, and obtaining abnormal information; determining the time attribute of the request according to the abnormal information, and establishing a corresponding Markov model; respectively establishing a current game theory model and a future game theory model, so as to obtain the current optimal solution and the future optimal solution, so as to adjust the network protection strategy. The accuracy of the analysis is improved, and the adaptability of the security protection strategy is guaranteed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of data analysis, and more specifically, to a method for analyzing abnormal DNS requests. Background Art

[0002] The abnormal DNS request resolution method involves multiple background technologies and comprehensive analysis methods. When analyzing abnormal DNS requests, the following are some key background technologies and comprehensive explanations:

[0003] DNS protocol and request resolution: DNS (Domain Name System) is a protocol used on the Internet to resolve domain names into IP addresses. Resolving DNS requests involves understanding the DNS protocol, including message structure, flags, query types, and response codes. This is the basis for resolving requests.

[0004] Threat intelligence and malicious domain name analysis: With threat intelligence and malicious domain name analysis, information about known malicious domain names, IP addresses, and attack patterns from different sources can be collected and integrated. This information can be used to compare and identify potential malicious behavior in abnormal DNS requests.

[0005] In the prior art, the resolution accuracy of abnormal DNS requests is poor, and misjudgment and missed judgment problems often occur, resulting in poor adaptability of network security protection strategies.

[0006] Therefore, how to improve parsing accuracy and strategy adaptability is a technical problem that needs to be solved. Summary of the invention

[0007] The present invention provides a method for parsing abnormal DNS requests, which is used to solve the technical problems of poor parsing accuracy and low policy adaptability in the prior art. The method comprises:

[0008] Obtain previous abnormal DNS requests and establish multiple abnormal standard indicators based on the logs of previous abnormal DNS requests;

[0009] Obtain the DNS request to be detected, and generate multiple abnormal indicators based on the log of the DNS request to be detected;

[0010] Compare the abnormal index with the abnormal standard index to determine whether the DNS request to be detected is an abnormal DNS request;

[0011] If the DNS request to be detected is an abnormal DNS request, the request message is analyzed and abnormal information is obtained;

[0012] Determine the time attributes of the request based on the anomaly information and build a corresponding Markov model to predict future DNS requests;

[0013] The current game theory model and the future game theory model are established respectively to obtain the current optimal solution and the future optimal solution to adjust the network protection strategy.

[0014] In some embodiments of the present application, multiple abnormal standard indicators are established based on the logs of previous abnormal DNS requests, including:

[0015] Obtain all characteristic data types in the logs of previous abnormal DNS requests;

[0016] Acquire threat intelligence data, and obtain features and attack patterns of abnormal DNS requests, thereby determining a first feature set;

[0017] Compare normal DNS request data with abnormal DNS request data to obtain difference feature data, and determine a second feature set;

[0018] The data features of normal and abnormal DNS requests are trained and classified through machine learning algorithms, characteristic data related to abnormal requests are identified, and a third characteristic set is determined;

[0019] The first feature set, the second feature set and the third feature set all include feature types and abnormal data ranges corresponding to the types;

[0020] A plurality of abnormal standard indicators are established based on the first feature set, the second feature set and the third feature set.

[0021] In some embodiments of the present application, multiple abnormal standard indicators are established based on the first feature set, the second feature set, and the third feature set, including:

[0022] The intersection feature data type of the first feature set, the second feature set and the third feature set is used as the first abnormal standard indicator type, and the union of the abnormal data ranges corresponding to the intersection feature data type of the first feature set, the second feature set and the third feature set is used as the data range of the first abnormal standard indicator type;

[0023] The intersection feature data types of the first feature set, the second feature set and the third feature set that intersect each other are used as the second abnormal standard indicator type, and the average value of the abnormal data range corresponding to the intersection feature data types of the first feature set, the second feature set and the third feature set that intersect each other is used as the data range of the second abnormal standard indicator type;

[0024] A plurality of abnormal standard indicators are established according to the first abnormal standard indicator type, the data range of the first abnormal standard indicator type, the second abnormal standard indicator type, and the data range of the second abnormal standard indicator type.

[0025] In some embodiments of the present application, the request message is analyzed and abnormal information is obtained, including:

[0026] Capture DNS request messages and parse DNS request message structures;

[0027] The message structure includes a message header, a query part, and a response part;

[0028] Analyze the message header to obtain the nature and characteristics of the request;

[0029] Analyze the query part to get the requested resource and query type;

[0030] Analyze the response part and get the data returned by the DNS server;

[0031] Check message fields and values ​​based on the nature, characteristics, resources and query type of the request and the data returned by the DNS server to obtain anomaly information.

[0032] In some embodiments of the present application, the time attribute of the request is determined according to the abnormal information, and a corresponding Markov model is established to predict future DNS requests, including:

[0033] Get the average time interval between multiple requests and define the time attributes as continuous time and discrete time;

[0034] ;

[0035] Among them, T is the average time interval, n is the number of requests - 1, is the i-th time interval, is an exponential function, for The minimum value, for The maximum value, is a preset constant;

[0036] A first Markov model is established for requests in continuous time, and a second Markov model is established for requests in discrete time;

[0037] Future DNS requests are predicted based on the first Markov model or the second Markov model.

[0038] In some embodiments of the present application, time attributes are defined as continuous time and discrete time, including:

[0039] If the number of times that the average time interval in the multiple requests is greater than the time interval exceeds a first preset number, the time attribute is defined as a continuous time;

[0040] If the number of times that the average time interval in the multiple requests is less than the time interval exceeds a second preset number, the time attribute is defined as discrete time.

[0041] In some embodiments of the present application, a current game theory model and a future game theory model are established respectively, including:

[0042] Identify the attackers and defenders in the game;

[0043] Obtain historical attack records and build the attacker's attack strategy based on the historical attack records;

[0044] Build the defender's defense strategy based on the defender's DNS request resolution method;

[0045] Define attack benefit, attack cost, defense benefit, and defense cost;

[0046] Information asymmetry rules are established, and there are preset differences between historical attack records and the actual attacker's attack strategies to ensure the attacker's advanced nature.

[0047] In some embodiments of the present application, the current optimal solution and the future optimal solution are obtained to adjust the network protection strategy, including:

[0048] Obtain Nash equilibrium solutions based on current game theory models and future game theory models;

[0049] Adjust network protection strategies based on Nash equilibrium solutions.

[0050] In some embodiments of the present application, adjusting the network protection strategy based on the Nash equilibrium solution includes:

[0051] If the number of Nash equilibrium solutions is 1, the network protection strategy is adjusted according to the Nash equilibrium solution;

[0052] Otherwise, compare the benefits of multiple Nash equilibrium solutions and adjust the network protection strategy according to the Nash equilibrium solution with the largest benefit.

[0053] By applying the above technical solution, previous abnormal DNS requests are obtained, and multiple abnormal standard indicators are established based on the logs of previous abnormal DNS requests; the DNS request to be detected is obtained, and multiple abnormal indicators are generated based on the logs of the DNS request to be detected; the abnormal indicators and abnormal standard indicators are compared to determine whether the DNS request to be detected is an abnormal DNS request; if the DNS request to be detected is an abnormal DNS request, the request message is analyzed and abnormal information is obtained; the time attribute of the request is determined based on the abnormal information, and a corresponding Markov model is established to predict future DNS requests; the current game theory model and the future game theory model are established respectively to obtain the current optimal solution and the future optimal solution to adjust the network protection strategy. This application improves the accuracy of the analysis by establishing multiple abnormal standard indicators based on the logs of previous abnormal DNS requests to determine whether the DNS request to be detected is abnormal. By establishing the current game theory model and the future game theory model to adjust the network protection strategy, the adaptability of the security protection strategy is guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0055] Figure 1 A schematic diagram of a process for resolving an abnormal DNS request proposed in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0057] The present application embodiment provides a method for resolving abnormal DNS requests, such as Figure 1 As shown, the method comprises the following steps:

[0058] Step S101, obtaining previous abnormal DNS requests, and establishing a plurality of abnormal standard indicators according to the logs of previous abnormal DNS requests.

[0059] In this embodiment, the abnormal standard index is a standard for judging whether a DNS request is abnormal, for example, a large number of query requests, an abnormal domain name length, etc.

[0060] In some embodiments of the present application, multiple abnormal standard indicators are established based on the logs of previous abnormal DNS requests, including:

[0061] Obtain all characteristic data types in the logs of previous abnormal DNS requests;

[0062] Acquire threat intelligence data, and obtain features and attack patterns of abnormal DNS requests, thereby determining a first feature set;

[0063] Compare normal DNS request data with abnormal DNS request data to obtain difference feature data, and determine a second feature set;

[0064] The data features of normal and abnormal DNS requests are trained and classified through machine learning algorithms, characteristic data related to abnormal requests are identified, and a third characteristic set is determined;

[0065] The first feature set, the second feature set and the third feature set all include feature types and abnormal data ranges corresponding to the types;

[0066] A plurality of abnormal standard indicators are established based on the first feature set, the second feature set and the third feature set.

[0067] In some embodiments of the present application, multiple abnormal standard indicators are established based on the first feature set, the second feature set, and the third feature set, including:

[0068] The intersection feature data type of the first feature set, the second feature set and the third feature set is used as the first abnormal standard indicator type, and the union of the abnormal data ranges corresponding to the intersection feature data type of the first feature set, the second feature set and the third feature set is used as the data range of the first abnormal standard indicator type;

[0069] The intersection feature data types of the first feature set, the second feature set and the third feature set that intersect each other are used as the second abnormal standard indicator type, and the average value of the abnormal data range corresponding to the intersection feature data types of the first feature set, the second feature set and the third feature set that intersect each other is used as the data range of the second abnormal standard indicator type;

[0070] A plurality of abnormal standard indicators are established according to the first abnormal standard indicator type, the data range of the first abnormal standard indicator type, the second abnormal standard indicator type, and the data range of the second abnormal standard indicator type.

[0071] Step S102, obtaining the DNS request to be detected, and generating multiple abnormal indicators according to the log of the DNS request to be detected.

[0072] In this embodiment, the type of abnormal indicator is the same as the abnormal standard indicator.

[0073] Step S103, comparing the abnormal index with the abnormal standard index, so as to determine whether the DNS request to be detected is an abnormal DNS request.

[0074] In this embodiment, by comparing the abnormal index and the abnormal standard index, it is detected whether the DNS request to be detected is an abnormal DNS request.

[0075] Step S104: if the DNS request to be detected is an abnormal DNS request, the request message is analyzed and abnormal information is obtained.

[0076] In this embodiment, once an abnormal DNS request is detected, the request message is further analyzed to obtain more information, which may involve checking the source IP address, target domain name, query type and other related fields, and obtaining the time information of the abnormal request.

[0077] In some embodiments of the present application, the request message is analyzed and abnormal information is obtained, including:

[0078] Capture DNS request messages and parse DNS request message structures;

[0079] The message structure includes a message header, a query part, and a response part;

[0080] Analyze the message header to obtain the nature and characteristics of the request;

[0081] Analyze the query part to get the requested resource and query type;

[0082] Analyze the response part and get the data returned by the DNS server;

[0083] Check message fields and values ​​based on the nature, characteristics, resources and query type of the request and the data returned by the DNS server to obtain anomaly information.

[0084] In this embodiment, the abnormality analysis includes the following steps:

[0085] Capture DNS request packets: Use a network packet capture tool (such as Wireshark) to capture DNS request packets on the network. This can help obtain actual network traffic data for analysis.

[0086] Parsing DNS request message structure DNS request messages follow a specific protocol structure, including a message header and a query part. Use parsing tools or programming languages ​​(such as Python's dnspython library) to parse and extract the various fields of the message.

[0087] Analyze the message header: The message header contains important information about the DNS request, such as identification number, flag bit, query number, response number, etc. Analyzing these fields can help you understand the nature and characteristics of the request.

[0088] Analyze the query part: The query part contains the queried domain name, query type (such as A record, MX record, etc.) and query category. This information can be used to identify the requested resource and specific query type.

[0089] Parse the response part (if any): If there is a response part, you need to parse the structure and fields of the response message to understand the data returned by the DNS server.

[0090] Check message fields and values: Checking the values ​​of various fields, such as flags and response codes, can help determine whether the DNS request is successful or not, and whether there are any anomalies or errors.

[0091] Step S105, determining the time attribute of the request according to the abnormal information, and establishing a corresponding Markov model, so as to predict future DNS requests.

[0092] In this embodiment, the time attribute is divided into continuous and discrete, and the Markov model can be divided into two main types: discrete-time Markov chain and continuous-time Markov process. Discrete-time Markov chain means that the system transitions from one state to another at discrete time points. At each time point, the state of the system can be determined according to the probability transition matrix. Continuous-time Markov process means that the system transitions from one state to another at continuous time. It uses transition rate to describe the speed and probability of state transition.

[0093] In some embodiments of the present application, the time attribute of the request is determined according to the abnormal information, and a corresponding Markov model is established to predict future DNS requests, including:

[0094] Get the average time interval between multiple requests and define the time attributes as continuous time and discrete time;

[0095] ;

[0096] Among them, T is the average time interval, n is the number of requests - 1, is the i-th time interval, is an exponential function, for The minimum value, for The maximum value, is a preset constant;

[0097] A first Markov model is established for requests in continuous time, and a second Markov model is established for requests in discrete time;

[0098] Future DNS requests are predicted based on the first Markov model or the second Markov model.

[0099] In this embodiment, Indicates the correction of the deviation to the mean.

[0100] In some embodiments of the present application, time attributes are defined as continuous time and discrete time, including:

[0101] If the number of times that the average time interval in the multiple requests is greater than the time interval exceeds a first preset number, the time attribute is defined as a continuous time;

[0102] If the number of times that the average time interval in the multiple requests is less than the time interval exceeds a second preset number, the time attribute is defined as discrete time.

[0103] Step S106, respectively establishing a current game theory model and a future game theory model, so as to obtain a current optimal solution and a future optimal solution to adjust the network protection strategy.

[0104] In some embodiments of the present application, a current game theory model and a future game theory model are established respectively, including:

[0105] Identify the attackers and defenders in the game;

[0106] Obtain historical attack records and build the attacker's attack strategy based on the historical attack records;

[0107] Build the defender's defense strategy based on the defender's DNS request resolution method;

[0108] Define attack benefit, attack cost, defense benefit, and defense cost;

[0109] Information asymmetry rules are established, and there are preset differences between historical attack records and the actual attacker's attack strategies to ensure the attacker's advanced nature.

[0110] In this embodiment, the following contents are specifically included:

[0111] Strategy selection:

[0112] Attacker strategy: Attackers may choose to use a large number of randomly generated malicious domain names to bypass detection by defense mechanisms. They may also use low-frequency queries and hiding techniques to make the attack more covert.

[0113] Defender strategy: Defenders can choose to use machine learning-based anomaly detection algorithms to identify abnormal DNS requests. They can monitor traffic patterns, domain name length, unusual query types, etc. to determine whether there is abnormal behavior.

[0114] Benefits and costs:

[0115] Attacker benefits: Attacker benefits may include successfully hijacking user traffic, stealing sensitive information, or conducting phishing activities. This may bring financial or other illegal benefits.

[0116] Attacker costs: The attacker's costs may involve acquiring or registering a large number of malicious domain names, developing and maintaining attack tools, and technical expenses for bypassing defense mechanisms.

[0117] Defender benefits: Defender benefits include protecting network security, preventing data breaches and business interruptions, and maintaining organizational reputation and trust.

[0118] Defender costs: Defender costs include investments in security equipment and tools, threat intelligence collection and analysis, training and education, etc.

[0119] Information Asymmetry:

[0120] Attacker advantages: The attacker may have advanced attack techniques, unknown malicious domains, or vulnerabilities that enable him to circumvent defense mechanisms and maintain information asymmetry.

[0121] Defender efforts: Defenders need to continuously acquire threat intelligence, participate in security communities, and share information to narrow the information asymmetry gap and update defense strategies and tools.

[0122] Strategic balance:

[0123] Attackers and defenders seek to achieve a strategic balance by constantly adjusting and improving their strategies. Attackers may try to change their attack methods, while defenders will improve their defense mechanisms based on new threat intelligence and technological advances.

[0124] In some embodiments of the present application, the current optimal solution and the future optimal solution are obtained to adjust the network protection strategy, including:

[0125] Obtain Nash equilibrium solutions based on current game theory models and future game theory models;

[0126] Adjust network protection strategies based on Nash equilibrium solutions.

[0127] In some embodiments of the present application, adjusting the network protection strategy based on the Nash equilibrium solution includes:

[0128] If the number of Nash equilibrium solutions is 1, the network protection strategy is adjusted according to the Nash equilibrium solution;

[0129] Otherwise, compare the benefits of multiple Nash equilibrium solutions and adjust the network protection strategy according to the Nash equilibrium solution with the largest benefit.

[0130] By applying the above technical solution, previous abnormal DNS requests are obtained, and multiple abnormal standard indicators are established based on the logs of previous abnormal DNS requests; the DNS request to be detected is obtained, and multiple abnormal indicators are generated based on the logs of the DNS request to be detected; the abnormal indicators and abnormal standard indicators are compared to determine whether the DNS request to be detected is an abnormal DNS request; if the DNS request to be detected is an abnormal DNS request, the request message is analyzed and abnormal information is obtained; the time attribute of the request is determined based on the abnormal information, and a corresponding Markov model is established to predict future DNS requests; the current game theory model and the future game theory model are established respectively to obtain the current optimal solution and the future optimal solution to adjust the network protection strategy. This application improves the accuracy of the analysis by establishing multiple abnormal standard indicators based on the logs of previous abnormal DNS requests to determine whether the DNS request to be detected is abnormal. By establishing the current game theory model and the future game theory model to adjust the network protection strategy, the adaptability of the security protection strategy is guaranteed.

[0131] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.

[0132] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for resolving abnormal DNS requests, characterized in that: The method comprises: Obtain previous abnormal DNS requests and establish multiple abnormal standard indicators based on the logs of previous abnormal DNS requests; Obtain the DNS request to be detected, and generate multiple abnormal indicators based on the log of the DNS request to be detected; Compare the abnormal index with the abnormal standard index to determine whether the DNS request to be detected is an abnormal DNS request; If the DNS request to be detected is an abnormal DNS request, the request message is analyzed and abnormal information is obtained; Determine the time attributes of the request based on the anomaly information and build a corresponding Markov model to predict future DNS requests; Establish the current game theory model and the future game theory model respectively, so as to obtain the current optimal solution and the future optimal solution to adjust the network protection strategy; Analyze the request message and obtain abnormal information, including: Capture DNS request messages and parse DNS request message structures; The message structure includes a message header, a query part, and a response part; Analyze the message header to obtain the nature and characteristics of the request; Analyze the query part to get the requested resource and query type; Analyze the response part and get the data returned by the DNS server; Check message fields and values ​​based on the nature, characteristics, resources and query type of the request and the data returned by the DNS server to obtain anomaly information; Determine the time attributes of the request based on the abnormal information and build the corresponding Markov model to predict future DNS requests, including: Get the average time interval between multiple requests and define the time attributes as continuous time and discrete time; ; Among them, T is the average time interval, n is the number of requests - 1, is the i-th time interval, is an exponential function, for The minimum value, for The maximum value, is a preset constant; A first Markov model is established for requests in continuous time, and a second Markov model is established for requests in discrete time; Predicting future DNS requests based on the first Markov model or the second Markov model; And define time attributes as continuous time and discrete time, including: If the number of times that the average time interval in the multiple requests is greater than the time interval exceeds a first preset number, the time attribute is defined as a continuous time; If the number of times that the average time interval in the multiple requests is less than the time interval exceeds a second preset number, the time attribute is defined as a discrete time; Establish the current game theory model and the future game theory model respectively, including: Identify the attackers and defenders in the game; Obtain historical attack records and build the attacker's attack strategy based on the historical attack records; Build the defender's defense strategy based on the defender's DNS request resolution method; Define attack benefit, attack cost, defense benefit, and defense cost; Establish information asymmetry rules, and there are preset differences between historical attack records and actual attackers' attack strategies to ensure the attackers' advanced attack strategies; This will help you get the current optimal solution and the future optimal solution to adjust your network protection strategy, including: Obtain Nash equilibrium solutions based on current game theory models and future game theory models; Adjust network protection strategies based on Nash equilibrium solutions.

2. The abnormal DNS request resolution method according to claim 1, characterized in that: Based on the logs of previous abnormal DNS requests, we have established several abnormal standard indicators, including: Obtain all characteristic data types in the logs of previous abnormal DNS requests; Acquire threat intelligence data, and obtain the characteristics and attack patterns of abnormal DNS requests, thereby determining a first feature set; Compare normal DNS request data with abnormal DNS request data to obtain difference feature data, and determine a second feature set; The data features of normal and abnormal DNS requests are trained and classified through machine learning algorithms, feature data related to abnormal requests are identified, and a third feature set is determined; The first feature set, the second feature set and the third feature set all include feature types and abnormal data ranges corresponding to the types; A plurality of abnormal standard indicators are established based on the first feature set, the second feature set and the third feature set.

3. The abnormal DNS request resolution method according to claim 2, characterized in that: Based on the first feature set, the second feature set and the third feature set, multiple abnormal standard indicators are established, including: The intersection feature data type of the first feature set, the second feature set and the third feature set is used as the first abnormal standard indicator type, and the union of the abnormal data ranges corresponding to the intersection feature data type of the first feature set, the second feature set and the third feature set is used as the data range of the first abnormal standard indicator type; The intersection feature data types of the first feature set, the second feature set and the third feature set that intersect each other are used as the second abnormal standard indicator type, and the average value of the abnormal data range corresponding to the intersection feature data types of the first feature set, the second feature set and the third feature set that intersect each other is used as the data range of the second abnormal standard indicator type; A plurality of abnormal standard indicators are established according to the first abnormal standard indicator type, the data range of the first abnormal standard indicator type, the second abnormal standard indicator type, and the data range of the second abnormal standard indicator type.

4. The abnormal DNS request resolution method according to claim 3, characterized in that: Adjust network protection strategies based on Nash equilibrium solutions, including: If the number of Nash equilibrium solutions is 1, the network protection strategy is adjusted according to the Nash equilibrium solution; Otherwise, compare the benefits of multiple Nash equilibrium solutions and adjust the network protection strategy according to the Nash equilibrium solution with the largest benefit.