Machine Learning-Based Electronic Device Authentication Methods and Systems

Abstract

This invention discloses a machine learning-based method and system for electronic device authentication. The method relates to the field of digital data processing technology and includes the following steps: collecting authentication key update delay data; quantifying and judging the authentication key update delay data; collecting authentication reliability data; and quantifying and judging the authentication reliability data. This invention quantifies the latency in the key update process using the acquired authentication key update delay data to determine whether optimization is needed. If so, authentication reliability data is obtained after key update delay optimization; otherwise, authentication reliability data is directly obtained. The obtained authentication reliability data is then used to quantify the reliability of the electronic device's authentication process to determine whether authentication reliability optimization is needed. This improves the authentication reliability of electronic devices and solves the problem of low authentication reliability in existing technologies due to key update delays and asynchrony.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data processing technology, and in particular to a machine learning-based method and system for electronic device authentication. Background Technology

[0002] With the rapid development of information technology, the information stored and processed by electronic devices is becoming increasingly sensitive, including personal identification information, financial data, and trade secrets. The leakage of this information can lead to serious consequences, such as financial loss, reputational damage, and loss of competitive advantage. Therefore, ensuring the secure storage and transmission of information has become crucial, and electronic devices, as carriers of information storage and processing, serve as the first line of defense for information security. By integrating various authentication technologies, electronic devices can effectively prevent unauthorized access, ensuring that only legitimate users can access sensitive information.

[0003] Existing technologies utilize minute differences generated during the manufacturing process of integrated circuits to generate unique keys. The PUF (Physically Unclonable Function) chip in electronic devices generates keys for each verification. Electronic devices, as nodes in a blockchain network, use blockchain technology to achieve decentralized identity verification.

[0004] For example, the invention patent application with publication number CN114065152A discloses an authentication method, authentication system, and electronic device, which includes: providing sub-character set information to the user sequentially according to a set number of times; obtaining unlock character set information input by the user based on the sub-character set information; retrieving stored preset character set information; comparing the unlock character set information with the preset character set information, and confirming successful unlocking when they are the same, and confirming unlocking failure when they are different.

[0005] For example, the invention patent announcement CN113139174B describes an authentication method and apparatus, which includes: displaying a first password on a first screen of an electronic device when authentication is required; receiving a first input from a user on a second screen of the electronic device, the content of which is a second password; and determining that authentication is successful in response to the first input if the second password matches the first password.

[0006] However, in the process of implementing the inventive technical solution in the embodiments of this application, it was found that the above-mentioned technology has at least the following technical problems:

[0007] In the application scenario of substation equipment access control, in the existing technology, when the request processing speed cannot keep up with the request arrival speed, the queue of requests waiting to be processed will continue to grow. Due to resource conflicts, the request processing time is extended, resulting in increased equipment response latency. Resources are occupied for a long time, and some requests may be dropped or fail to be processed. After the equipment receives the key update instruction, due to CPU occupation and network latency, the key update cannot be completed in time. Errors occur during the key update process, resulting in update failure. The equipment still uses the old key or cannot work normally. The key update progress of different equipment is not synchronized. The hardware configuration and processing capabilities of equipment of different voltage levels are different, resulting in differences in key update processing speed. There is a problem of low reliability of electronic equipment authentication due to key update delay and asynchrony. Summary of the Invention

[0008] This application provides a machine learning-based electronic device authentication method and system, which solves the problem of low reliability of electronic device authentication caused by key update delays and asynchrony in the prior art, and improves the reliability of electronic device authentication.

[0009] This application provides a machine learning-based electronic device authentication method, including the following steps: During key update, acquiring authentication key update delay data, where key update refers to updating the key used for authentication before the electronic device in the preset power system begins authentication; quantifying the latency in the key update process using the authentication key update delay data to obtain an authentication key update delay efficiency index; determining whether to perform key update delay optimization based on the obtained authentication key update delay efficiency index, where key update delay optimization refers to improving the real-time performance of the key update process through dynamic resource adjustment and CPU time slice adjustment of the key update task; if key update delay optimization is performed, acquiring authentication reliability data after key update delay optimization; otherwise, directly acquiring authentication reliability data; quantifying the reliability of the electronic device's authentication process using the authentication reliability data to obtain an authentication reliability efficiency index; determining whether to perform authentication reliability optimization based on the obtained authentication reliability efficiency index; if yes, performing authentication reliability optimization, where authentication reliability optimization refers to improving the reliability of the authentication process through dynamic adjustment of key update frequency and dynamic adjustment of multi-authentication; otherwise, directly performing authentication feedback.

[0010] This application provides an electronic device authentication system based on machine learning. The machine learning-based electronic device authentication method includes an authentication key update delay data acquisition module, an authentication key update delay data quantification and judgment module, an authentication reliability data acquisition module, and an authentication reliability data quantification and judgment module: The authentication key update delay data acquisition module acquires authentication key update delay data during the key update process; the authentication key update delay data quantification and judgment module quantifies the delay in the key update process using the authentication key update delay data to obtain an authentication key update delay efficiency index, and determines whether to perform key update delay optimization based on the obtained authentication key update delay efficiency index; the authentication reliability data acquisition module acquires authentication reliability data after key update delay optimization if optimization is required, otherwise directly acquires authentication reliability data; the authentication reliability data quantification and judgment module quantifies the reliability in the electronic device's authentication process using the authentication reliability data to obtain an authentication reliability efficiency index, and determines whether to perform authentication reliability optimization based on the obtained authentication reliability efficiency index, and if so, performs authentication reliability optimization.

[0011] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0012] 1. By quantifying the latency in the key update process using the obtained authentication key update latency data, it is determined whether optimization is required. If optimization is required, authentication reliability data is obtained after key update latency optimization; otherwise, it is obtained directly. The reliability of the electronic device's authentication process is quantified using the obtained authentication reliability data to determine whether authentication reliability optimization is required. This achieves the effectiveness of key update and synchronization, thereby improving the authentication reliability of electronic devices and solving the problem of low authentication reliability of electronic devices caused by key update latency and asynchrony in existing technologies.

[0013] 2. By processing the authentication key update delay data and the delay in the key update process, an authentication key update delay performance index is obtained, thereby achieving accurate quantification of the delay in the key update process. The authentication key update delay performance index is then used to determine whether to dynamically adjust resources, thus optimizing the key update synchronization mechanism, reducing authentication failures caused by asynchrony, thereby improving authentication security, and achieving a smooth transition in the key update process, reducing operation interruptions caused by delay.

[0014] 3. By processing the authentication reliability data and the reliability of the electronic device authentication process, an authentication reliability performance index is obtained, thereby achieving accurate quantification of the reliability in the authentication process. The authentication reliability performance index is then used to determine whether to optimize the authentication reliability. The key update synchronization mechanism is optimized. By dynamically adjusting the key update frequency and multi-authentication, failures caused by outdated keys or single authentication are reduced, thereby improving the accuracy of authentication and thus improving the reliability of electronic device authentication. Attached Figure Description

[0015] Figure 1 A flowchart of a machine learning-based electronic device authentication method provided in this application embodiment;

[0016] Figure 2 A schematic diagram of the structure of an electronic device authentication system based on machine learning provided in an embodiment of this application;

[0017] Figure 3 A flowchart illustrating the dynamic resource adjustment process of the machine learning-based electronic device authentication method provided in this application embodiment;

[0018] Figure 4 A flowchart illustrating the dynamic adjustment of the key update frequency in the machine learning-based electronic device authentication method provided in this application embodiment. Detailed Implementation

[0019] This application provides a machine learning-based electronic device authentication method and system, which solves the problem of low authentication reliability of electronic devices caused by key update delays and asynchrony in the prior art. By obtaining authentication key update delay data and quantifying the delay in the key update process, an authentication key update delay performance index is obtained to determine whether optimization is needed. If optimization is needed, authentication reliability data is obtained after key update delay optimization; otherwise, it is obtained directly. Based on the obtained authentication reliability performance index, it is determined whether authentication reliability optimization is needed, thereby improving the authentication reliability of electronic devices.

[0020] The technical solution in this application embodiment addresses the aforementioned issue of low reliability in electronic device authentication due to key update delays and asynchrony. The overall approach is as follows:

[0021] The authentication key update delay efficiency index is obtained by quantifying the latency of the key update process using authentication key update delay data acquired during the key update process. Then, based on the obtained authentication key update delay efficiency index, it is determined whether to optimize the key update delay. If key update delay optimization is performed, authentication reliability data is obtained after the optimization; otherwise, authentication reliability data is obtained directly. Finally, the reliability of the electronic device's authentication process is quantified using the authentication reliability data to obtain the authentication reliability efficiency index, and it is determined whether to optimize the authentication reliability, thereby improving the authentication reliability of the electronic device.

[0022] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0023] like Figure 1 The diagram shows a flowchart of a machine learning-based electronic device authentication method provided in this application embodiment. This method is applied to a machine learning-based electronic device authentication system and includes the following steps: During key update, acquiring authentication key update delay data, where key update refers to updating the key used for authentication before the electronic device in the preset power system begins authentication; quantifying the latency in the key update process using the authentication key update delay data to obtain an authentication key update delay efficiency index; determining whether to perform key update delay optimization based on the obtained authentication key update delay efficiency index, where key update delay optimization refers to improving the real-time performance of the key update process through dynamic resource adjustment and CPU time slice adjustment of the key update task; if key update delay optimization is performed, acquiring authentication reliability data after key update delay optimization; otherwise, directly acquiring authentication reliability data; quantifying the reliability of the electronic device's authentication process using the authentication reliability data to obtain an authentication reliability efficiency index; determining whether to perform authentication reliability optimization based on the obtained authentication reliability efficiency index; if yes, performing authentication reliability optimization, where authentication reliability optimization refers to improving the reliability of the authentication process through dynamic adjustment of key update frequency and dynamic adjustment of multi-authentication; otherwise, directly performing authentication feedback.

[0024] In this embodiment, during the key update process, anomaly detection of key usage behavior is performed using unsupervised learning algorithms such as Isolation Forest in machine learning, thereby triggering the key update. Because the request processing speed cannot keep up with the request arrival speed, the request processing time is prolonged, leading to increased response latency in the electronic device, which in turn affects the subsequent real-time analysis of the key update. Due to the low reliability of the electronic device's authentication process, errors occur during authentication, resulting in authentication failure. In this application, key update latency optimization includes dynamic resource adjustment and CPU time slice adjustment for the key update task. Real-time performance during the key update process is improved by dynamically adjusting resources and the CPU time slice of the key update task. Authentication reliability optimization includes dynamic adjustment of the key update frequency and dynamic adjustment of multi-factor authentication. The reliability of the authentication process is improved by dynamically adjusting the key update frequency and multi-factor authentication.

[0025] Specifically, the authentication key update latency data includes key update response time, number of concurrent encryption requests, number of key update task priority preemption attempts, and amount of encrypted data per unit time. It should be added that obtaining the authentication key update latency performance index previously included obtaining the authentication key update latency threshold and authentication key update latency correction amount from the built authentication database. The authentication key update latency threshold includes the key update response time threshold, standard value for the number of concurrent encryption requests, standard value for the number of key update task priority preemption attempts, and standard value for the amount of encrypted data per unit time. The authentication key update latency correction amount includes corrections for key update response time, number of concurrent encryption requests, number of key update task priority preemption attempts, and amount of encrypted data per unit time.

[0026] Specifically, the authentication reliability data includes authentication success rate, key update success rate, authentication key update delay efficiency index, and time synchronization failure rate. It should be noted that obtaining the authentication reliability efficiency index previously included retrieving authentication reliability thresholds and authentication reliability corrections from the built authentication database. The authentication reliability thresholds include authentication success rate thresholds, key update success rate thresholds, authentication key update delay thresholds, and time synchronization failure rate thresholds. The authentication reliability corrections include authentication success rate corrections, key update success rate corrections, authentication key update delay corrections, and time synchronization failure rate corrections.

[0027] It should be added that the aforementioned authentication database is a database established before the design of the machine learning-based electronic device authentication method to store various settings data. The database includes, but is not limited to, the set key update response time, the set number of concurrent encryption requests, the set number of key update task priority preemption times, the set amount of encrypted data per unit time, and the set key update success rate threshold. These values ​​are directly set by technical personnel. The set key update success rate threshold can be determined based on the actual application scenario of electronic device authentication. For example, the set key update success rate threshold can be represented by the average of the historical key update success rates in the historical database within a set monitoring frequency range, as determined by pre-set staff. In addition, the various values ​​in the database can be set and fine-tuned by technical personnel based on actual debugging.

[0028] like Figure 2 The diagram shows the structure of an electronic device authentication system based on machine learning provided in this application embodiment. The system includes: an authentication key update delay data acquisition module, an authentication key update delay data quantification and judgment module, an authentication reliability data acquisition module, and an authentication reliability data quantification and judgment module. The authentication key update delay data acquisition module acquires authentication key update delay data during the key update process. The authentication key update delay data quantification and judgment module quantifies the delay in the key update process using the authentication key update delay data to obtain an authentication key update delay efficiency index, and determines whether to perform key update delay optimization based on the obtained efficiency index. The authentication reliability data acquisition module acquires authentication reliability data after key update delay optimization if optimization is performed; otherwise, it directly acquires authentication reliability data. The authentication reliability data quantification and judgment module quantifies the reliability in the electronic device's authentication process using the authentication reliability data to obtain an authentication reliability efficiency index, and determines whether to perform authentication reliability optimization based on the obtained efficiency index; if so, it performs authentication reliability optimization.

[0029] In this embodiment, the electronic device includes, but is not limited to, relay protection devices, remote terminal units, etc. When the relay protection device needs to communicate with other devices, it will trigger a key update process. For example, when a relay protection device needs to report fault information to the control center, it uses a new session key to encrypt the data. During the key update process, the start and end times of the key update operation are recorded, uniquely identifying the device performing the key update and indicating the type of key being updated, such as a session key or a long-term key. When maintenance personnel remotely log in to the relay protection device to modify settings, the relay protection device, upon receiving the request, generates a random number and uses its own private key to digitally sign the combination of the random number sent by the workstation, the random number generated by the relay protection device, and the timestamp sent by the workstation. The signature result, public key, and the random number generated by the relay protection device are returned to the workstation. The workstation uses the public key to verify whether the signature is correct. If correct, the relay protection device is confirmed as authentic; otherwise, it is deemed authentic. If the relay protection device is authentic, the workstation uses the public key to encrypt the user password, administrator information, and a new timestamp, and sends it to the relay protection device. The relay protection device uses its private key to decrypt the data, verifies whether the password matches the database record, and checks whether the user's permissions allow the operation (e.g., non-administrators cannot modify settings).

[0030] Furthermore, the specific steps to obtain the authentication key update latency efficiency index are as follows: First, correct the analysis results of the ratio of key update response time to key update response time threshold using a key update response time correction factor, denoted as the key update response time impact value; second, correct the analysis results of the ratio of the deviation of the number of concurrent encryption requests to the standard value of the number of concurrent encryption requests using a concurrent encryption request number correction factor, denoted as the concurrent encryption request number impact value; third, correct the analysis results of the ratio of the deviation of the number of key update task priority preemption attempts to the standard value of the number of key update task priority preemption attempts using a key update task priority preemption attempt correction factor, denoted as the key update task priority preemption attempt... The impact value of the number of times; the ratio of the deviation of the encrypted data volume per unit time to the standard value of the encrypted data volume per unit time is corrected by the correction of the encrypted data volume per unit time, and recorded as the impact value of the encrypted data volume per unit time; the impact value of the key update response time, the impact value of the number of concurrent encryption requests, the impact value of the number of times the key update task priority is preempted, and the impact value of the encrypted data volume per unit time are coupled to obtain the authentication key update delay efficiency index; the authentication key update delay efficiency index represents the quantitative data of the degree of influence of the key update response time, the number of concurrent encryption requests, the number of times the key update task priority is preempted, and the encrypted data volume per unit time on the real-time performance of key updates.

[0031] It should be added that the deviation in the number of concurrent encryption requests represents the absolute value of the difference between the number of encryption requests sent and the standard value; the deviation in the number of times the key update task priority preemption occurs represents the absolute value of the difference between the number of times the key update task priority preemption occurs and the standard value; and the deviation in the amount of data encrypted per unit time represents the absolute value of the difference between the amount of data encrypted per unit time and the standard value.

[0032] In this embodiment, the specific formula for obtaining the authentication key update delay performance index is as follows:

[0033]

[0034] α1+α2+α3+α4=1;

[0035] ξ represents the efficiency index of authentication key update delay.

[0036] T represents the key update response time, which is the time required from triggering the key update request to the actual completion of the key update, obtained through the log recorder.

[0037] T0 represents the key update response time threshold, which is obtained from the authentication database.

[0038] Q represents the number of concurrent encryption requests, which refers to the number of encryption operations processed simultaneously. A dedicated queue monitoring tool, such as Apache Kafka, is used to record the number of encryption requests being processed to obtain the number of concurrent encryption requests.

[0039] Q0 represents the standard value for the number of concurrent encrypted requests, obtained from the authentication database.

[0040] Z represents the number of times the key update task priority preemption occurred. It refers to the number of times the key update task was interrupted by a low-priority task during task scheduling. A low-priority task is a task with a lower priority than the key update task. The number of times the key update task priority preemption occurred is obtained by displaying the task scheduling status and priority in real time through the operating system scheduler log.

[0041] Z0 represents the standard value for the number of times the key update task priority preemption can be obtained from the authentication database.

[0042] S represents the amount of encrypted data per unit time, which refers to the amount of encrypted data that the CPU can process per unit time. The amount of encrypted data per unit time is obtained by monitoring the CPU performance through the CPU performance counter.

[0043] S0 represents the standard value of the amount of encrypted data per unit time, obtained from the authentication database.

[0044] α1 is the correction amount for the key update response time obtained from the authentication database.

[0045] α2 is a correction factor for the number of concurrent encrypted requests retrieved from the authentication database.

[0046] α3 is the adjustment amount for the priority preemption count of the key update task obtained from the authentication database.

[0047] α4 is the correction amount for the amount of encrypted data per unit time obtained from the authentication database.

[0048] In this example, the key update response time correction, concurrent encryption request count correction, key update task priority preemption count correction, and encrypted data volume per unit time correction represent the degree of impact of the key update response time, concurrent encryption request count, key update task priority preemption count, and encrypted data volume per unit time on the authentication key update delay, respectively, as preset in the authentication database. Specifically, the authentication database stores preset correction values ​​corresponding to the key update response time, concurrent encryption request count, key update task priority preemption count, and encrypted data volume per unit time. These correction values ​​have a pre-defined mapping relationship with the key update response time, concurrent encryption request count, key update task priority preemption count, and encrypted data volume per unit time. This mapping relationship can be one-to-one or many-to-one. For example, in practical applications, the real-time key update response time, concurrent encryption request count, key update task priority preemption count, and encrypted data volume per unit time can be input into this mapping relationship to quickly obtain the corresponding correction values.

[0049] In this example, the values ​​of the key update response time correction, concurrent encryption request number correction, key update task priority preemption number correction, and unit time encrypted data volume correction are typically between 0 and 1, and the sum of the four is 1.

[0050] In this embodiment, the parameters involved in obtaining the authentication key update delay performance index are not independent of each other, but rather influence each other.

[0051] Specifically, the larger the number of concurrent encryption requests, the more system resources such as CPU will be occupied by encryption requests. The key update task needs to wait for these resources to become available or compete for resources with other tasks, resulting in a longer key update response time. The more times the key update task priority is preempted, the more time is required for state saving and restoration. In addition, the preemption and restoration process will also introduce additional overhead such as context switching, which will increase the total execution time of the key update task, that is, the key update response time will be longer. The increase in the amount of encrypted data per unit time means that the overall system load is increased, and the key update response time will be longer.

[0052] There is a positive correlation between key update response time and the authentication key update latency efficiency index. A longer key update response time indicates a slower key update process, a slower system response time, and a higher authentication key update latency efficiency index. Similarly, there is a positive correlation between the absolute value of the difference between the number of concurrent encryption requests and the standard value of the number of concurrent encryption requests, and the authentication key update latency efficiency index. A larger absolute value of this difference indicates excessive system load, leading to increased resource contention, and a higher authentication key update latency efficiency index. Furthermore, the absolute value of the difference between the number of times the key update task priority preemption occurs and the standard value of the authentication key update latency efficiency index is also positively correlated with the authentication key update latency efficiency index. The key update latency efficiency index shows a positive correlation. The larger the absolute value of the difference between the number of times the key update task priority is preempted and the standard value of the number of times the key update task priority is preempted, the more frequently the key update task is interrupted, requiring more time to switch tasks, and the higher the authentication key update latency efficiency index. The absolute value of the difference between the amount of encrypted data per unit time and the standard value of the amount of encrypted data per unit time also shows a positive correlation with the authentication key update latency efficiency index. The larger the absolute value of the difference between the amount of encrypted data per unit time and the standard value of the amount of encrypted data per unit time, the more the system needs to process a large amount of encrypted data, which will increase the system load and intensify resource competition, resulting in a higher authentication key update latency efficiency index.

[0053] By analyzing the correlation between key update response time, number of concurrent encryption requests, number of key update task priority preemption times, and amount of encrypted data per unit time with the authentication key update latency efficiency index, the latency in the key update process can be quantified more accurately. Based on the quantification results, key update latency can be optimized more promptly and accurately, thereby improving authentication security and key update synchronization, thus enhancing authentication security.

[0054] Furthermore, the obtained authentication key update delay efficiency index determines whether key update delay optimization should be performed. The specific steps are as follows: if the authentication key update delay efficiency index is less than or equal to the preset authentication key update delay security threshold, then no key update delay optimization is performed; if the authentication key update delay efficiency index is within the preset authentication key update delay security range, then the key update time determines whether to perform dynamic resource adjustment. If so, then after performing dynamic resource adjustment, it is determined whether to adjust the CPU time slice of the key update task; otherwise, it is directly determined whether to adjust the CPU time slice of the key update task. The preset authentication key update delay security range indicates that the authentication key update delay efficiency index is greater than the preset authentication key update delay security threshold obtained from the authentication database and less than the range corresponding to the preset authentication key update delay threshold; if the authentication key update delay efficiency index is greater than or equal to the preset authentication key update delay threshold, then an authentication delay prompt is sent to the preset administrator.

[0055] As a further solution, the specific steps for determining whether to dynamically adjust resources based on key update time are as follows: If the key update time is less than or equal to a preset key update time security threshold, it indicates that the system response is rapid, but there is still a deviation. The initial key version drift rate is compensated by the obtained key version drift rate compensation amount to improve the timeliness of key updates. The key version drift rate compensation amount is obtained by mapping the minimum key update deviation time and the authentication key update delay efficiency index into the authentication database. The minimum key update deviation time represents the difference between the key update time and the preset key update time security threshold. If the key update time is within the preset key update time security range, no dynamic resource adjustment is performed, and the system is maintained. The current key version drift rate, and the preset key update time security range, represent the range where the key update time is greater than the preset key update time security threshold obtained from the authentication database but less than the corresponding preset key update time threshold. If the key update time is greater than or equal to the preset key update time threshold, it indicates that it will lead to resource conflicts or synchronization failures. The initial key version drift rate is corrected by the obtained key version drift rate correction amount to reduce resource scheduling conflicts. The key version drift rate correction amount is obtained by mapping the maximum key update deviation time and the authentication key update delay efficiency index into the authentication database. The maximum key update deviation time represents the difference between the key update time and the preset key update time threshold.

[0056] In this embodiment, as Figure 3The diagram shows a flowchart of the resource dynamic adjustment process of the machine learning-based electronic device authentication method provided in this application embodiment. During the key update process, it is determined whether the key update time is less than or equal to a preset key update time security threshold. If so, the key version drift rate compensation amount is calculated, and the initial key version drift rate is compensated by the key version drift rate compensation amount. Otherwise, it is determined whether the key update time is within the preset key update time security interval. If so, the current configuration is maintained. Otherwise, the key version drift rate correction amount is calculated, and the initial key version drift rate is corrected by the key version drift rate correction amount.

[0057] In the key update process of power system electronic equipment, the key version drift rate is dynamically adjusted by comparing the key update time with a preset key update time security threshold to improve the timeliness of key updates and reduce resource scheduling conflicts. The obtained key version drift rate compensation amount is used to compensate for the initial key version drift rate to improve the timeliness of key updates. This compensation amount is added to the initial key version drift rate to compensate for time deviations and ensure key update synchronization. The obtained key version drift rate correction amount is used to correct the initial key version drift rate to reduce resource scheduling conflicts. Subtracting the correction amount from the initial key version drift rate reduces the CPU, network, and other resource consumption of the key update task, thus avoiding scheduling conflicts.

[0058] Further, the specific steps for determining whether to adjust the CPU time slice of the key update task are as follows: If dynamic resource adjustment has been performed, the newly acquired authentication key update latency efficiency index is recorded as the authentication key update latency efficiency index to be corrected; otherwise, the current authentication key update latency efficiency index is recorded as the authentication key update latency efficiency index to be corrected. If the average key operation latency is less than or equal to the preset average key operation latency safety threshold, it indicates that the current CPU resource allocation meets the real-time requirements, and there is no need to adjust the CPU time slice of the key update task. If the average key operation latency is within the preset average key operation latency safety range, the initial reserved CPU time slice of the key update task is adjusted by the obtained increase in the reserved CPU time slice of the key update task to improve the utilization of CPU resources. The increase in the reserved CPU time slice of the key update task is achieved by adjusting the average key operation latency. The average delay and the performance index of the authentication key update delay to be corrected are mapped into the authentication database. The preset average key operation delay security range represents the range where the average key operation delay is greater than the preset average key operation delay security threshold obtained from the authentication database and less than the corresponding preset average key operation delay threshold. If the average key operation delay is greater than or equal to the preset average key operation delay threshold, the CPU time slice of the initial key update task is compensated by the obtained CPU time slice compensation amount of the key update task to improve the timeliness of the key update task. The CPU time slice compensation amount of the key update task is obtained by mapping the maximum deviation of the average key operation delay and the performance index of the authentication key update delay to be corrected into the authentication database. The maximum deviation of the average key operation delay represents the difference between the average key operation delay and the preset average key operation delay threshold.

[0059] In this embodiment, key updates take precedence over status detection. Key updates have the highest priority and cannot be preempted. Status detection can be interrupted or downgraded by key updates. Basic CPU time slices are allocated to tasks of different priorities to ensure the timely execution of high-priority tasks. The system adjusts the CPU time slices for key update tasks based on a comparison between the average latency of key operations (i.e., the average latency of key update operations in the historical database) and a preset average latency threshold for key operations.

[0060] The initial reserved CPU time slice for the key update task is adjusted by increasing the reserved CPU time slice of the key update task to improve CPU resource utilization. This increase is then added to the initial reserved CPU time slice to ensure that the key update task has higher execution priority in CPU resource contention. The initial CPU time slice for the key update task is compensated by increasing the CPU time slice of the key update task to improve its timeliness. This compensation is added to the initial CPU time slice, thereby prioritizing the execution of the key update task through a preemptive scheduling mechanism and shortening the response latency of the key update.

[0061] Furthermore, the specific process for obtaining the authentication reliability performance index is as follows: The analysis results of the ratio of authentication success rate to authentication success rate threshold are corrected using the authentication success rate correction factor, and this is recorded as the effective value of the authentication success rate; the analysis results of the ratio of key update success rate to key update success rate threshold are corrected using the key update success rate correction factor, and this is recorded as the effective value of the key update success rate; the analysis results of the ratio of authentication key update delay threshold to the authentication key update delay performance index are corrected using the authentication key update delay correction factor, and this is recorded as the effective value of the authentication key update delay; the analysis results of the ratio of time synchronization failure rate threshold to time synchronization failure rate are corrected using the time synchronization failure rate correction factor, and this is recorded as the effective value of the time synchronization failure rate; the effective values ​​of authentication success rate, key update success rate, authentication key update delay, and time synchronization failure rate are coupled to obtain the authentication reliability performance index; the authentication reliability performance index represents the quantitative data on the degree of influence of the combined effects of authentication success rate, key update success rate, authentication key update delay, and time synchronization failure rate on authentication reliability.

[0062] In this embodiment, the specific formula for obtaining the authentication reliability performance index is as follows:

[0063]

[0064] ρ1+ρ2+ρ3+ρ4=1;

[0065] ψ represents the reliability and effectiveness index of identity verification.

[0066] Y represents the authentication success rate, which is the ratio of the number of successful authentications to the total number of authentications. The authentication success rate is obtained through log records.

[0067] Y0 represents the authentication success rate threshold, obtained from the authentication database.

[0068] G represents the key update success rate, which is the ratio of the number of successfully completed key updates to the total number of key updates. The key update success rate is obtained through a counter.

[0069] G0 represents the key update success rate threshold, obtained from the authentication database.

[0070] ε0 represents the authentication key update delay threshold, which is obtained from the authentication database.

[0071] ε represents the efficiency index of the authentication key update delay to be compared. Specifically, if key update delay optimization has been performed, the efficiency index of the authentication key update delay to be compared represents the efficiency index of the authentication key update delay after key update delay optimization. It should be noted that if key update delay optimization has not been performed, the efficiency index of the authentication key update delay is the same as the efficiency index of the authentication key update delay to be compared.

[0072] B0 represents the time synchronization failure rate threshold, obtained from the authentication database.

[0073] B represents the time synchronization failure rate, which is the ratio of the number of times a time synchronization operation fails to complete to the total number of attempts. The time synchronization failure rate is obtained through NTP / PTP monitoring.

[0074] The k1 constant term is a number set to avoid points in the data that are meaningless.

[0075] ρ1 is the authentication success rate correction amount obtained from the authentication database.

[0076] ρ2 is the key update success rate correction amount obtained from the authentication database.

[0077] ρ3 is the amount of delay correction for updating the authentication key obtained from the authentication database.

[0078] ρ4 is the time synchronization failure rate correction amount obtained from the authentication database.

[0079] In this example, the authentication success rate correction, key update success rate correction, authentication key update delay correction, and time synchronization failure rate correction represent the degree of impact of the authentication success rate, key update success rate, authentication key update delay efficiency index, and time synchronization failure rate—pre-defined in the authentication database—on authentication reliability. Specifically, the authentication database stores preset correction values ​​corresponding to the authentication success rate, key update success rate, authentication key update delay efficiency index, and time synchronization failure rate. These correction values ​​have a pre-defined mapping relationship with the authentication success rate, key update success rate, authentication key update delay efficiency index, and time synchronization failure rate. This mapping relationship can be one-to-one or many-to-one. For example, in practical applications, real-time authentication success rate, key update success rate, authentication key update delay efficiency index, and time synchronization failure rate can be input into this mapping relationship to quickly obtain the corresponding correction values.

[0080] In this example, the values ​​of the authentication success rate correction, key update success rate correction, authentication key update delay correction, and time synchronization failure rate correction are typically between 0 and 1, and the sum of the four is 1.

[0081] In this embodiment, the parameters involved in obtaining the authentication reliability performance index are not independent of each other, but rather influence each other.

[0082] Specifically, a higher authentication success rate means a faster authentication process and a smaller latency efficiency index for the authentication key update; a higher key update success rate ensures the continuity and efficiency of encrypted communication and a smaller latency efficiency index for the authentication key update; a higher time synchronization failure rate can cause problems in the authentication and key update processes, resulting in a smaller latency efficiency index for the authentication key update and a smaller key update success rate.

[0083] There is a positive correlation between authentication success rate and authentication reliability performance index. A higher authentication success rate indicates a more reliable authentication process with fewer failures, resulting in a higher authentication reliability performance index. Similarly, there is a positive correlation between key update success rate and authentication reliability performance index. A higher key update success rate helps maintain the continuous reliability and security of authentication, resulting in a higher authentication reliability performance index. Conversely, there is a negative correlation between the pending authentication key update delay performance index and authentication reliability performance index. A higher pending authentication key update delay performance index means greater key update delay and lower efficiency, resulting in a lower authentication reliability performance index. Finally, there is a negative correlation between time synchronization failure rate and authentication reliability performance index. A higher time synchronization failure rate leads to authentication failures or key update failures, resulting in a lower authentication reliability performance index.

[0084] By analyzing the correlation between authentication success rate, key update success rate, authentication key update delay efficiency index, time synchronization failure rate and authentication reliability efficiency index, the reliability of the authentication process can be quantified more accurately. Based on the quantification results, authentication reliability can be optimized more timely and accurately, thereby improving authentication accuracy and ultimately enhancing the reliability of electronic device authentication.

[0085] Furthermore, the specific steps for determining whether to perform authentication reliability optimization based on the obtained authentication reliability performance index are as follows: If the authentication data reliability performance index is greater than or equal to the preset authentication data reliability threshold obtained from the authentication database, it indicates that the current authentication reliability meets the requirements, and no authentication data reliability optimization adjustment is performed; if the authentication data reliability performance index is less than the preset authentication data reliability threshold obtained from the authentication database, authentication data reliability optimization adjustment is performed, and the re-obtained authentication data reliability performance index is recorded as the authentication data reliability performance index to be compared; if the authentication data reliability performance index to be compared is greater than or equal to the preset authentication data reliability threshold, the electronic device password is entered; otherwise, a warning prompt is triggered; authentication data reliability optimization adjustment includes dynamic adjustment of key update frequency and dynamic adjustment of multi-authentication; the specific steps for dynamic adjustment of key update frequency are as follows: if the clock synchronization interval is lower than or equal to the preset clock synchronization interval security threshold, the initial clock synchronization frequency is compensated by the obtained clock synchronization frequency increase to reduce the impact of clock synchronization frequency fluctuations. The synchronization deviation caused by clock drift is determined by the increase in clock synchronization frequency. This is obtained by mapping the minimum clock synchronization interval deviation to the authentication data reliability performance index in the authentication database. The minimum clock synchronization interval deviation represents the difference between the clock synchronization interval and the preset clock synchronization interval security threshold. If the clock synchronization interval is within the preset clock synchronization interval security range, the initial key update frequency is compensated by the obtained increase in key update frequency to improve the timeliness of key updates. The increase in key update frequency is obtained by mapping the clock synchronization interval to the authentication data reliability performance index in the authentication database. The preset clock synchronization interval security range represents the range where the clock synchronization interval is greater than the preset clock synchronization interval security threshold obtained from the authentication database but less than the corresponding preset clock synchronization interval threshold. If the clock synchronization interval is greater than or equal to the preset clock synchronization interval threshold, it indicates a serious clock synchronization problem. In this case, adjusting the key update frequency cannot solve the problem, so the system will stop dynamically adjusting the key update frequency and issue a key update rejection prompt to avoid authentication failure due to clock asynchrony.

[0086] In this embodiment, as Figure 4The flowchart shown is a process for dynamically adjusting the key update frequency in the machine learning-based electronic device authentication method provided in this application embodiment. It determines whether the clock synchronization interval is less than or equal to a preset clock synchronization interval security threshold. If so, it calculates the increase in clock synchronization frequency and compensates the initial clock synchronization frequency using this increase. Otherwise, it determines whether the clock synchronization interval is within the preset clock synchronization interval security range. If so, it calculates the increase in key update frequency and compensates the initial key update frequency using this increase. Otherwise, it sends a rejection message.

[0087] In the authentication process of electronic equipment in power systems, specifically: the initial clock synchronization frequency is compensated by the obtained increase in clock synchronization frequency to reduce synchronization deviation caused by clock drift. The increased clock synchronization frequency is added to the initial clock synchronization frequency to further reduce synchronization deviation caused by clock drift. Similarly, the initial key update frequency is compensated by the obtained increase in key update frequency to improve the timeliness of key updates. The increased key update frequency is added to the initial key update frequency to further improve the timeliness of key updates.

[0088] Furthermore, the specific steps for dynamic adjustment of multi-factor authentication are as follows: If the key update frequency is dynamically adjusted, the reliability efficiency index of the newly acquired authentication data is recorded as the reliability efficiency index of the authentication data to be corrected; otherwise, the current reliability efficiency index of the authentication data is recorded as the reliability efficiency index of the authentication data to be corrected. If the authentication time delay is less than or equal to the preset authentication time delay security threshold, the initial number of authentication factors is compensated by the increase in the number of authentication factors to improve authentication security. The increase in the number of authentication factors is obtained by mapping the minimum authentication time delay and the reliability efficiency index of the authentication data to be corrected into the authentication database. The minimum authentication time delay represents the difference between the authentication time delay and the preset authentication time delay security threshold. If the authentication time delay is within the preset authentication time delay security range, no dynamic adjustment of multi-factor authentication will be performed. The preset authentication time delay security range means that the authentication time delay is greater than the preset authentication time delay security threshold obtained from the authentication database and less than the range corresponding to the preset authentication time delay threshold. If the authentication time delay is greater than or equal to the preset authentication time delay threshold, the initial number of authentication factors will be corrected by the obtained verification factor number correction amount to improve authentication security. The verification factor number correction amount is obtained by mapping the maximum authentication time delay and the authentication data reliability performance index to be corrected into the authentication database. The maximum authentication time delay represents the difference between the authentication time delay and the preset authentication time delay threshold.

[0089] In this embodiment, in the authentication process of power system electronic equipment, the reliability performance index of the authentication data to be corrected is determined based on whether the key update frequency has been dynamically adjusted: if the key update frequency has been dynamically optimized, the performance index re-collected after optimization is used as the value to be corrected; if the adjustment has not been performed, the performance index calculated in real time is used directly as the value to be corrected. Multiple authentication is dynamically adjusted based on the comparison between the authentication time delay (the average time required to complete one authentication) and the preset authentication time delay threshold to improve the real-time performance of authentication. The initial number of authentication factors is compensated for by the increase in the number of authentication factors to improve authentication security. The increase in the number of authentication factors is added to the initial number of authentication factors to improve authentication security. If the initial authentication factors are password and hardware token and the mapped increase in the number of authentication factors is 1, the initial authentication factors are adjusted to password, hardware token, and fingerprint recognition, thus increasing security by adding authentication dimensions. The initial number of authentication factors is corrected by the correction amount of the number of authentication factors to improve authentication security. The initial number of authentication factors is subtracted from the correction amount to reduce authentication delay, while retaining core authentication factors to maintain basic security.

[0090] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0091] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0092] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0093] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0094] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0095] 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 machine learning-based electronic device authentication method, characterized in that, Includes the following steps: During the key update process, authentication key update delay data is obtained. The key update refers to updating the key used for authentication before the electronic equipment of the preset power system begins authentication. The latency of the key update process is quantified by the authentication key update latency data to obtain the authentication key update latency efficiency index. Based on the obtained authentication key update latency efficiency index, it is determined whether to perform key update latency optimization. The key update latency optimization means improving the real-time performance of the key update process by dynamically adjusting resources and adjusting the CPU time slice of the key update task. If key update delay optimization is performed, the authentication reliability data is obtained after the key update delay optimization is performed; otherwise, the authentication reliability data is obtained directly. The reliability of the electronic device's authentication process is quantified by authentication reliability data to obtain an authentication reliability performance index. Based on the obtained authentication reliability performance index, it is determined whether to perform authentication reliability optimization. If so, authentication reliability optimization is performed. The authentication reliability optimization means improving the reliability of the authentication process by dynamically adjusting the key update frequency and dynamically adjusting the multi-authentication. Otherwise, authentication feedback is directly performed. The specific steps for obtaining the authentication key update delay performance index are as follows: The analysis results of the ratio of key update response time to key update response time threshold are corrected by the key update response time correction amount, and this is recorded as the key update response time impact value. The result of the analysis of the proportion of the deviation of the number of concurrent encrypted requests to the standard value of the number of concurrent encrypted requests is corrected by the correction amount of the number of concurrent encrypted requests, and is recorded as the impact value of the number of concurrent encrypted requests. The result of the analysis of the ratio of the deviation of the number of times the priority of the key update task is corrected by the correction amount of the number of times the priority of the key update task is corrected to the standard value of the number of times the priority of the key update task is corrected, and is recorded as the impact value of the number of times the priority of the key update task is corrected. The correction is made by the ratio of the deviation of the encrypted data volume per unit time to the standard value of the encrypted data volume per unit time, and is recorded as the influence value of the encrypted data volume per unit time. The impact values ​​of key update response time, concurrent encryption request quantity, key update task priority preemption times, and encrypted data volume per unit time are coupled to obtain the authentication key update latency performance index. The authentication key update delay efficiency index represents a quantitative data representation of the combined impact of key update response time, number of concurrent encryption requests, number of key update task priority preemption attempts, and amount of encrypted data per unit time on the real-time performance of key updates.

2. The electronic device authentication method based on machine learning as described in claim 1, characterized in that, The authentication key update delay data includes key update response time, number of concurrent encryption requests, number of key update task priority preemption attempts, and amount of encrypted data per unit time; The process of obtaining the authentication key update delay performance index also includes obtaining the authentication key update delay threshold and the authentication key update delay correction amount from the constructed authentication database. The authentication key update delay threshold includes a key update response time threshold, a standard value for the number of concurrent encryption requests, a standard value for the number of times the key update task priority is preempted, and a standard value for the amount of encrypted data per unit time. The authentication key update delay correction includes the key update response time correction, the number of concurrent encryption requests correction, the key update task priority preemption count correction, and the amount of encrypted data per unit time correction. The authentication reliability data includes authentication success rate, key update success rate, authentication key update delay efficiency index to be compared, and time synchronization failure rate. The process of obtaining the authentication reliability performance index also includes obtaining the authentication reliability threshold and authentication reliability correction amount from the constructed authentication database. The authentication reliability thresholds include the authentication success rate threshold, the key update success rate threshold, the authentication key update delay threshold, and the time synchronization failure rate threshold. The authentication reliability correction includes authentication success rate correction, key update success rate correction, authentication key update delay correction, and time synchronization failure rate correction.

3. The electronic device authentication method based on machine learning as described in claim 1, characterized in that, The specific steps for determining whether to perform key update delay optimization based on the obtained authentication key update delay performance index are as follows: If the authentication key update delay performance index is less than or equal to the preset authentication key update delay security threshold, then no key update delay optimization will be performed; If the authentication key update delay efficiency index is within the preset authentication key update delay security range, then the key update time determines whether to perform dynamic resource adjustment. If so, after performing dynamic resource adjustment, it determines whether to adjust the CPU time slice of the key update task. Otherwise, it directly determines whether to adjust the CPU time slice of the key update task. The preset authentication key update delay security range means that the authentication key update delay efficiency index is greater than the preset authentication key update delay security threshold obtained from the authentication database and less than the range corresponding to the preset authentication key update delay threshold. If the authentication key update delay efficiency index is greater than or equal to the preset authentication key update delay threshold, an authentication delay notification will be sent to the preset administrator.

4. The electronic device authentication method based on machine learning as described in claim 3, characterized in that, The specific steps for determining whether to dynamically adjust resources based on the key update time are as follows: If the key update time is less than or equal to the preset key update time security threshold, the initial key version drift rate is compensated by the obtained key version drift rate compensation amount to improve the timeliness of key update. The key version drift rate compensation amount is obtained by mapping the minimum deviation time of key update and the authentication key update delay efficiency index into the authentication database. If the key update time is within the preset key update time security interval, no dynamic resource adjustment will be performed. The preset key update time security interval means that the key update time is greater than the preset key update time security threshold obtained from the authentication database and less than the range corresponding to the preset key update time threshold. If the key update time is greater than or equal to the preset key update time threshold, the initial key version drift rate is corrected by the obtained key version drift rate correction amount to reduce resource scheduling conflicts. The key version drift rate correction amount is obtained by mapping the maximum key update deviation time and the authentication key update delay efficiency index into the authentication database.

5. The machine learning-based electronic device authentication method as described in claim 4, characterized in that, The specific steps for determining whether to adjust the CPU time slice for the key update task are as follows: If dynamic resource adjustments are made, the newly acquired authentication key update delay performance index will be recorded as the authentication key update delay performance index to be corrected; otherwise, the current authentication key update delay performance index will be recorded as the authentication key update delay performance index to be corrected. If the average latency of key operations is less than or equal to the preset average latency security threshold for key operations, then the CPU time slice adjustment for the key update task will not be performed. If the average key operation latency is within the preset safe range of average key operation latency, the initial reserved CPU time slice of the key update task is adjusted by increasing the reserved CPU time slice of the key update task to improve the utilization of CPU resources. The increase in the reserved CPU time slice of the key update task is obtained by mapping the average key operation latency and the performance index of the authentication key update latency to be corrected into the authentication database. The preset safe range of average key operation latency means that the average key operation latency is greater than the preset safe threshold of average key operation latency obtained from the authentication database and less than the range corresponding to the preset safe threshold of average key operation latency. If the average delay of key operations is greater than or equal to the preset average delay threshold of key operations, the CPU time slice of the initial key update task is compensated by the obtained CPU time slice compensation amount of the key update task to improve the timeliness of the key update task. The CPU time slice compensation amount of the key update task is obtained by mapping the maximum deviation of the average delay of key operations and the performance index of the delay of the authentication key update to be corrected into the authentication database.

6. The machine learning-based electronic device authentication method as described in claim 2, characterized in that, The specific process for obtaining the identity verification reliability performance index is as follows: The results of the analysis of the ratio of the authentication success rate to the authentication success rate threshold are corrected by the authentication success rate correction amount, and recorded as the effective value of the authentication success rate. The result of the analysis of the ratio between the key update success rate and the key update success rate threshold is corrected by the key update success rate correction amount, and it is recorded as the effective value of the key update success rate. The results of the analysis of the proportion of the authentication key update delay threshold to the authentication key update delay efficiency index of the comparison pair are corrected by the authentication key update delay correction amount, and this is recorded as the effective value of the authentication key update delay. The time synchronization failure rate threshold and the ratio of time synchronization failure rate are corrected by the time synchronization failure rate correction amount, and the result is recorded as the effective value of time synchronization failure rate. By coupling the valid values ​​of authentication success rate, key update success rate, authentication key update delay and time synchronization failure rate, the authentication reliability performance index is obtained. The authentication reliability performance index represents a quantitative data representation of the combined impact of authentication success rate, key update success rate, authentication key update delay, and time synchronization failure rate on authentication reliability.

7. The machine learning-based electronic device authentication method as described in claim 1, characterized in that, The specific steps for determining whether to perform authentication reliability optimization based on the obtained authentication reliability performance index are as follows: If the reliability performance index of the authentication data is greater than or equal to the preset reliability threshold of the authentication data obtained from the authentication database, no optimization adjustment of the authentication data reliability will be performed. If the reliability performance index of the authentication data is less than the preset reliability threshold of the authentication data obtained from the authentication database, the reliability of the authentication data will be optimized and adjusted, and the reliability performance index of the newly obtained authentication data will be recorded as the reliability performance index of the authentication data to be compared. If the reliability efficiency index of the identity verification data to be compared is greater than or equal to the preset identity verification data reliability threshold, then the electronic device password is entered; otherwise, a warning prompt is triggered. The optimization and adjustment of the reliability of the authentication data includes dynamic adjustment of the key update frequency and dynamic adjustment of multi-factor authentication. The specific steps for dynamically adjusting the key update frequency are as follows: If the clock synchronization interval is lower than or equal to the preset clock synchronization interval safety threshold, the initial clock synchronization frequency is compensated by the obtained increase in clock synchronization frequency to reduce the synchronization deviation caused by clock drift. The increase in clock synchronization frequency is obtained by mapping the minimum deviation of clock synchronization interval to the authentication data reliability performance index into the authentication database. If the clock synchronization interval is within the preset clock synchronization interval safety range, the initial key update frequency is compensated by the obtained key update frequency increase to improve the timeliness of key updates. The key update frequency increase is obtained by mapping the clock synchronization interval and the authentication data reliability performance index into the authentication database. The preset clock synchronization interval safety range means that the clock synchronization interval is greater than the preset clock synchronization interval safety threshold obtained from the authentication database and less than the range corresponding to the preset clock synchronization interval threshold. If the clock synchronization interval is greater than or equal to the preset clock synchronization interval threshold, the key update frequency will not be dynamically adjusted, and a key update rejection prompt will be issued.

8. The machine learning-based electronic device authentication method as described in claim 7, characterized in that, The specific steps for dynamically adjusting the multi-factor authentication are as follows: If the key update frequency is dynamically adjusted, the reliability performance index of the newly acquired authentication data will be recorded as the reliability performance index of the authentication data to be corrected; otherwise, the current reliability performance index of the authentication data will be recorded as the reliability performance index of the authentication data to be corrected. If the authentication time delay is less than or equal to the preset authentication time delay security threshold, the initial number of authentication factors is compensated by the increase in the number of authentication factors to improve authentication security. The increase in the number of authentication factors is obtained by mapping the minimum authentication time delay and the reliability performance index of the authentication data to be corrected into the authentication database. If the authentication time delay is within the preset authentication time delay security range, no multi-authentication dynamic adjustment will be performed. The preset authentication time delay security range means that the authentication time delay is greater than the preset authentication time delay security threshold obtained from the authentication database and less than the range corresponding to the preset authentication time delay threshold. If the authentication time delay is greater than or equal to the preset authentication time delay threshold, the initial number of authentication factors is corrected by the obtained correction amount of the number of authentication factors to improve authentication security. The correction amount of the number of authentication factors is obtained by mapping the maximum authentication time delay and the reliability performance index of the authentication data to be corrected into the authentication database.

9. A machine learning-based electronic device authentication system, employing the machine learning-based electronic device authentication method as described in any one of claims 1-8, characterized in that, It includes an authentication key update delay data acquisition module, an authentication key update delay data quantification and judgment module, an authentication reliability data acquisition module, and an authentication reliability data quantification and judgment module. The authentication key update delay data acquisition module is used to acquire authentication key update delay data during the key update process. The authentication key update delay data quantification and judgment module is used to obtain the authentication key update delay efficiency index by quantifying the delay in the key update process through authentication key update delay data, and to determine whether to perform key update delay optimization based on the obtained authentication key update delay efficiency index; The authentication reliability data acquisition module is used to acquire authentication reliability data after key update delay optimization if key update delay optimization is performed; otherwise, it directly acquires authentication reliability data. The authentication reliability data quantification and judgment module is used to quantify the reliability of the electronic device's authentication process through authentication reliability data to obtain an authentication reliability performance index. Based on the obtained authentication reliability performance index, it determines whether to perform authentication reliability optimization. If so, it performs authentication reliability optimization.

Images