Algorithm optimization method and system for Internet of Things equipment

Abstract

The invention provides an algorithm optimization method and system for Internet of Things equipment, and relates to the technical field of Internet of Things, and the method comprises the steps: constructing a basic framework of an Internet of Things equipment algorithm through employing a TFF algorithm, and embedding an encryption region in the TFF algorithm framework; converting the original data into an encryption quantitative index according to a Laplacian mechanism; establishing an association mechanism between the security parameter and the self-optimization parameter according to the encryption quantitative index to obtain risk assessment data; according to the risk assessment data, dynamically adjusting safety parameters in the self-optimization process of the Internet of Things algorithm to obtain dynamically adjusted safety parameters; a safety auditing system is built according to the dynamically adjusted safety parameters, and early warning is carried out when the algorithm execution time exceeds a threshold value. According to the method, the data is quantized in combination with the differential privacy Laplacian mechanism, so that efficient self-optimization of the Internet of Things algorithm is guaranteed, the data security of the Internet of Things equipment is enhanced, and complex and changeable data security challenges are effectively coped with.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to an algorithm optimization method for Internet of Things devices and an algorithm optimization system for Internet of Things devices. Background Art

[0002] With the rapid development of IoT technology, a large number of IoT devices are being widely used in various fields, such as smart homes, industrial automation, and intelligent transportation. During operation, these IoT devices need to continuously collect, transmit, and process massive amounts of data. The performance of their built-in algorithms directly affects key indicators such as the overall device performance, energy consumption, and data service quality.

[0003] Existing IoT device algorithms often have numerous shortcomings. In industrial automation scenarios, some traditional data analysis algorithms use complex machine learning models to analyze equipment operating parameters collected by numerous sensors on the production line for fault prediction or quality monitoring. These models are inherently computationally intensive, requiring IoT devices to possess significant computing power. However, after deployment, most IoT device algorithms typically operate according to initially set rules and parameters, making it difficult to automatically optimize and adjust to changing environmental conditions and data characteristics. Some IoT device algorithms may be built with a focus on functional implementation and performance optimization, with insufficient consideration given to data security. Summary of the Invention

[0004] The present invention provides an algorithm optimization method and system for an Internet of Things device, which are used to solve the defects of the existing Internet of Things device algorithm in the lack of dynamic self-optimization capability and poor data security.

[0005] In one aspect, the present invention provides an algorithm optimization method for an IoT device, wherein the TFF algorithm is selected as the basic framework of the IoT device algorithm, and the input data of the IoT device is raw data, including: S1: Use the TFF algorithm to build the basic framework of the IoT device algorithm, embed the encrypted area into the TFF algorithm framework, and obtain the encrypted TFF algorithm framework.

[0006] S2: The input data of the IoT device is raw data. According to the encrypted TFF algorithm framework and combined with the Laplace mechanism, the raw data is converted into encrypted quantitative indicators.

[0007] S3: Based on the encryption quantitative indicators, the association mechanism between security parameters and self-optimization parameters is established to obtain risk assessment data.

[0008] S4: Dynamically adjust security parameters during the self-optimization process of the IoT algorithm according to the risk assessment data to obtain dynamically adjusted security parameters.

[0009] S5: Build a security audit system based on dynamically adjusted security parameters and analyze behavioral profiles. When the algorithm execution time exceeds the threshold, the IoT device will issue a synchronous warning, obtaining an IoT device algorithm with dynamic self-optimization capabilities and data security.

[0010] According to the algorithm optimization method for IoT devices provided by the present invention, in step S1, the TFF algorithm framework includes: a layered architecture of the algorithm framework, with the underlying data processing and basic operations as the foundation layer. The middle layer is set as the core algorithm layer. The upper layer is planned as the security and self-optimization management layer.

[0011] According to an algorithm optimization method for an IoT device provided by the present invention, in step S1, the specific steps of constructing a TFF algorithm framework for data encryption include: S11: Build the overall architecture of the secure feedback channel based on the characteristics of the TFF algorithm framework for data encryption and business requirements.

[0012] S12: Integrate an encryption communication protocol in the feedback channel and use a symmetric encryption algorithm to encrypt the transmitted data.

[0013] S13: Use a combination of digital certificates and biometrics to verify the identity of users and devices.

[0014] According to an algorithm optimization method for an IoT device provided by the present invention, in step S2, the Laplace mechanism is used to protect privacy information, and the specific steps of converting raw data into quantifiable data are as follows: S21: Allocating privacy budget to Laplace mechanism.

[0015] S22: Add noise and perturb the original data according to the privacy budget to obtain hidden data.

[0016] S23: Classify and standardize the hidden data.

[0017] S24: Construct a mapping relationship between the standardized hidden data and quantitative indicators.

[0018] S25: Aggregate and analyze the hidden data according to the mapping relationship to obtain encrypted quantitative indicators.

[0019] According to an algorithm optimization method for an IoT device provided by the present invention, in step S3, the encryption quantitative index analyzes the mutual influence relationship between the security parameters and the self-optimization parameters, and the specific steps of establishing the association mechanism are as follows: S31: Extract and analyze the features of encryption quantitative indicators, sort out the correlation performance of security parameters and self-optimization parameters in each link of data processing, and obtain preliminary correlation data.

[0020] S32: Construct a logical relationship chain of mutual influence between security parameters and self-optimization parameters to obtain detailed correlation data.

[0021] S33: Based on the detailed correlation data and historical data, corresponding security parameters and self-optimization parameter adjustment rules and strategies under different impact levels are formulated to obtain a correlation mechanism.

[0022] S34: Apply the correlation mechanism to simulate the impact of parameter changes on the overall security status of the data under different security scenarios to obtain risk assessment data.

[0023] According to an algorithm optimization method for an IoT device provided by the present invention, in step S32, the specific steps of constructing a relationship chain are: S321: Sort out the complete path of mutual influence between safety parameters and self-optimization parameters.

[0024] S322: Draw a logical relationship diagram of the mutual influence between the safety parameters and the self-optimization parameters through visualization.

[0025] S323: Input the quantitative model based on the complete path and the logical relationship diagram, substitute different parameter value ranges, and obtain detailed correlation data.

[0026] According to an algorithm optimization method for an IoT device provided by the present invention, in step S4, the specific steps of dynamically adjusting the security parameters are: S41: Collect security data and algorithm operation status in real time, conduct quantitative assessment of data security risks, and obtain dynamic risk scores.

[0027] S42: Adjust the security parameters through dynamic risk scoring, and calculate the self-optimization parameter compensation value to obtain the security self-optimization parameter.

[0028] S43: Verify the impact index of security self-optimization parameters on the performance and security of IoT device algorithms in a simulation environment.

[0029] S44: Optimize and improve the safety parameter adjustment plan based on the impact index, re-determine the safety parameter adjustment range, and obtain dynamically adjusted safety parameters.

[0030] According to an algorithm optimization method for an IoT device provided by the present invention, in step S42, the specific steps of adjusting security parameters through dynamic risk scoring are as follows: S421: Use neural network algorithm to establish a mapping model.

[0031] S422: Obtain the dynamic risk score of the current system and input it into the mapping model. Calculate the value to be adjusted for each security parameter according to the mapping model to obtain the security parameter value to be adjusted.

[0032] S423: Use support vector machine to establish an association model.

[0033] S424: Obtain a safety optimization parameter based on the influence of the value of the safety parameter to be adjusted on the self-optimization parameter, input the safety optimization parameter into the association model to calculate the value of the self-optimization parameter to be adjusted, and obtain the safety self-optimization parameter.

[0034] According to an algorithm optimization method for an IoT device provided by the present invention, building a security audit system includes: S51: Use the stream processing engine to parse and extract features from the original data stream to generate real-time behavior profiles.

[0035] S52: Based on the risk assessment results and behavioral baseline, a dynamic threshold model is used to determine the real-time monitoring threshold.

[0036] S53: Use sliding window technology to perform weighted aggregation calculations on behavioral features within the past N cycles, combined with anomaly detection algorithms to identify behavioral patterns that deviate from normal distribution. When the algorithm execution time exceeds the threshold, a message is synchronously pushed to the IoT device terminal.

[0037] S54: Analyze the risk transmission path triggered by abnormal regional tracing behavior, associate the early warning and disposal results with the dynamic threshold optimization strategy, and form a closed-loop management and control system from behavior monitoring to strategy iteration.

[0038] On the other hand, the present invention also provides an algorithm optimization system for an Internet of Things device, comprising: Algorithm encryption module: embeds the encryption area into the TFF algorithm framework to obtain the encrypted TFF algorithm framework.

[0039] Data quantification module: Based on the encrypted TFF algorithm framework and combined with the Laplace mechanism, the original data is converted into quantifiable data to obtain encrypted quantitative indicators.

[0040] Risk assessment module: Analyzes the mutual influence between security parameters and self-optimization parameters based on encrypted quantitative indicators to obtain data risk assessment.

[0041] Dynamic adjustment module: Dynamically adjusts security parameters during the self-optimization process of the IoT algorithm based on data risk assessment to obtain dynamically adjusted security parameters.

[0042] Synchronous early warning module: Build a security audit system based on dynamically adjusted security parameters. The security audit system analyzes behavioral profiles, synchronizes early warnings, and associates policy iterations for closed-loop control.

[0043] The algorithm optimization method and system for IoT devices provided by this invention deeply embeds the encryption mechanism into the TFF algorithm framework, combines it with the differential privacy Laplace mechanism to quantize data, constructs a security-self-optimization parameter association mechanism, dynamically adjusts security parameters, and establishes a closed-loop security audit system. These methods and systems address the core issues faced by IoT algorithms during data transmission, storage, and self-optimization, such as privacy leakage risks, rigid parameter configuration, and delayed security responses. The beneficial effects achieved are: This invention embeds the encrypted area into the TFF framework, converts the original data into encrypted quantitative indicators through the Laplace mechanism, and maintains data availability through quantization processing while protecting data privacy. It resolves the contradiction between privacy protection and algorithm optimization, enables the IoT algorithm to achieve efficient self-optimization in the data encryption state, and improves the robustness of the algorithm in complex environments.

[0044] The present invention establishes a security parameter and self-optimization parameter association mechanism based on encryption quantitative indicators, so that data risk assessment can quantitatively reflect the algorithm operation status. The dynamically adjusted security parameters can adjust the security policy in real time according to the assessment results, such as key rotation cycle, encryption strength, etc., forming a closed-loop management and control from risk identification to policy optimization, upgrading traditional static protection to an adaptive dynamic defense system, and significantly reducing the probability of IoT devices being attacked.

[0045] The security audit system of the present invention analyzes real-time behavioral portraits, triggers early warnings, and associates policy iterations to achieve risk transmission path tracing and dynamic threshold optimization, upgrading security management from passive response to active prevention. For example, when the algorithm execution time is abnormal, the system can automatically adjust the encryption level and push the repair strategy to the terminal, forming an intelligent decision-making chain of "monitoring-early warning-handling-optimization", reducing operation and maintenance costs while improving the overall security resilience of the Internet of Things system.

[0046] The present invention as a whole ensures efficient self-optimization of IoT algorithms while accurately assessing data risks, dynamically adapting security parameters, monitoring abnormal behaviors in real time and issuing timely warnings. It forms a closed-loop control through strategy iteration, greatly enhancing the data security of IoT systems and the stability of algorithm operation, effectively responding to complex and changing data security challenges, and facilitating the reliable development of IoT services. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0048] Figure 1This is a flow chart of an algorithm optimization method for an Internet of Things device provided by an embodiment of the present invention; Figure 2 This is a module diagram of an algorithm optimization system for an Internet of Things device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0050] The following combination Figure 1-Figure 2 The present invention describes an algorithm optimization method and system for an Internet of Things device.

[0051] like Figure 1 As shown, an embodiment of the present invention provides an algorithm optimization method for an Internet of Things device, the method comprising: S1: Select an algorithm framework that is both scalable and secure as the foundation. The algorithm framework's layered architecture prioritizes data processing and basic operations as the foundational layer. The middle layer is designated the core algorithm layer, which carries specific IoT business algorithms. The upper layer is designed as the security and self-optimization management layer. Clearly define interface specifications between layers and between regions to ensure secure and orderly data transmission between regions and facilitate subsequent expansion and adjustment.

[0052] A data encryption area is embedded in the framework, and the advanced encryption standard encryption algorithm is used to encrypt the transmitted data in real time. The asymmetric encryption algorithm is used to manage the key to ensure the confidentiality of the data during transmission and prevent data theft. At the same time, a strict two-way authentication mechanism is established. Through a combination of digital certificates, pre-shared keys and other methods, the legitimacy of IoT devices accessing the network and users accessing device data is verified to ensure the legitimacy of device and user access, and to protect the security of data at the source and during transmission.

[0053] S11: Build the overall architecture of the secure feedback channel based on the characteristics of the TFF algorithm framework for data encryption and business requirements. Define the various components of the channel, including the data sender, data receiver, and intermediate data transmission links. Ensure the architecture is highly secure and reliable, capable of resisting common cyberattacks and data leakage risks.

[0054] S12: Integrate an encrypted communication protocol into the feedback channel and use the data encryption area within the TFF framework to encrypt the transmitted data. Select an appropriate encryption algorithm, such as symmetric or asymmetric, and ensure the secure management of encryption keys. This encrypted communication protocol ensures the confidentiality and integrity of data during transmission, preventing data theft or tampering.

[0055] S13: To ensure that only legitimate users and devices can participate in the transmission of feedback data, a strict identity authentication and authorization mechanism is established. Multi-factor authentication methods, such as username / password, digital certificates, and biometrics, are used to verify the identities of users and devices. Furthermore, authorization management is implemented for access and transmission of feedback data based on user roles and permissions to prevent unauthorized access and data misuse.

[0056] S2: Based on the encrypted TFF algorithm framework and combined with the Laplace mechanism to protect privacy information, the original data is converted into a quantifiable signal to obtain an encrypted quantitative indicator.

[0057] S21: Use the Laplace mechanism to select an appropriate differential privacy model based on the characteristics of business feedback data and privacy protection requirements. Allocate a reasonable privacy budget for the differential privacy mechanism. The privacy budget is a parameter that measures the degree of privacy protection and controls the degree of noise added to the data. Determine the privacy budget for each user or each data transmission based on business requirements and data security risk assessment. Properly allocating the privacy budget can minimize the impact on data availability while ensuring privacy protection.

[0058] S22: Noise and perturb the original data according to the privacy budget to obtain hidden data. By adding an appropriate amount of noise to the data, attackers cannot accurately infer the private information of individual users from the data. At the same time, ensure that the added noise does not significantly affect the overall statistical characteristics of the data to ensure the effectiveness of subsequent data analysis and processing.

[0059] S23: Classify and standardize the actual business feedback data after differential privacy protection. Divide the feedback data into different categories based on different business dimensions and indicators, such as user satisfaction, system performance indicators, and business process efficiency. Standardize the data in each category to a unified format and scope to facilitate subsequent quantitative analysis.

[0060] S24: Define corresponding quantitative indicators for each category and establish a mapping relationship between feedback data and quantitative indicators. For example, for user satisfaction feedback, you can define a quantitative indicator as a satisfaction score, mapping the user's text feedback or selected options to a specific score value. Quantitative indicators can be used to transform complex business feedback into numerical signals that can be mathematically calculated and compared.

[0061] S25: Aggregate and analyze the converted quantitative signals. Based on business needs and analysis objectives, select an appropriate aggregation method, such as summation, average, or median, to aggregate the quantitative signals of different users or time periods. By analyzing the aggregated signals, patterns and trends are discovered, providing valuable information for self-optimization adjustments of the algorithm.

[0062] S3: Analyze the mutual influence between security parameters and self-optimization parameters. Security parameters include encryption strength, authentication frequency, learning rate, and update threshold. This allows the security parameters to be dynamically adjusted according to the data security risk assessment during the algorithm self-optimization process. At the same time, the impact on the self-optimization effect is also considered when adjusting the security policy.

[0063] S31: Extract and analyze the features of encryption quantitative indicators, sort out the correlation performance of security parameters and self-optimization parameters in each link of data processing, and obtain preliminary correlation data.

[0064] Using data processing techniques and analytical tools, we extract valuable features from cryptographic quantitative indicators. For example, features reflecting data transmission include the frequency, volume, and time intervals of data transmission. Data characteristics may include the numerical range, degree of dispersion, and distribution of data types. Algorithm-related data features focus on computational time during algorithm execution and trends in intermediate results. After acquiring these features, we delve deeper into how security parameters and self-optimization parameters interrelate at various stages of data processing. Security parameters typically encompass factors such as encryption strength, authentication frequency, and the granularity of access control. Self-optimization parameters primarily encompass aspects such as learning rate and update thresholds.

[0065] S32: Construct a logical relationship chain of mutual influence between security parameters and self-optimization parameters to obtain detailed correlation data.

[0066] S321: Integrate the influence of the safety parameters on the self-optimization parameters and the influence of the self-optimization parameters on the safety parameters, and sort out the complete path of mutual influence between the various parameters.

[0067] S322: Draw a logical relationship diagram of the mutual influence between safety parameters and self-optimization parameters through visualization, using nodes to represent each parameter, arrows to indicate the direction of influence, and annotate the corresponding quantitative relationship expression (such as the formula obtained from the previous modeling) on ​​the arrows to make the entire relationship chain more intuitive and clear.

[0068] S323: Based on the constructed logical relationship chain and the corresponding quantitative model, different parameter value ranges are substituted to derive the specific correlation data between the safety parameters and the self-optimization parameters under different circumstances.

[0069] The calculation formula of the learning rate in the self-optimization parameter is expressed as: ; Where ξ0 is the initial learning rate, α is a constant, and △E is the change in encryption strength.

[0070] S33: Based on the detailed correlation and historical data, the corresponding security parameter and self-optimization parameter adjustment rules and strategies under different levels of impact are formulated to obtain the correlation mechanism. Looking back at the large amount of historical data recorded by the IoT system during its past operation, these data cover various aspects of information such as data processing, security incidents, and the effectiveness of algorithm self-optimization in different periods and business scenarios. By mining and analyzing these historical data, it is possible to summarize the actual impact of changes in security parameters and self-optimization parameters on the overall system in various practical situations. For example, there have been cases where improper adjustment of encryption strength has caused the algorithm to converge too slowly and the risk of data leakage has increased. Or during certain business peak hours, the failure to adapt the authentication frequency in a timely manner has caused problems such as data transmission interruption and a decrease in the accuracy of algorithm prediction.

[0071] S34: Apply the correlation mechanism model to simulate the impact of parameter changes on the overall data security status under different security scenarios to obtain a data risk assessment. Design and construct a variety of simulation scenarios based on various actual security threats and common risk scenarios that IoT systems may face. For example, simulate external hacker attack scenarios, setting attack methods and frequencies of varying intensity to observe how the correlation mechanism drives adjustments to security parameters and self-optimization parameters in these situations, and the effectiveness of these adjustments in protecting data security. Simulate internal data leakage risk scenarios by manually setting different levels of permission abuse and data misoperation to test whether the correlation mechanism can promptly detect and reduce the possibility of data leakage through parameter adjustments. Simulate unstable data transmission scenarios caused by network failures, equipment failures, and other reasons to observe how the correlation mechanism responds to such abnormal situations to ensure data integrity and security.

[0072] S4: Build a security audit system to record security-related operations such as data access and encryption, and monitor abnormal behavior in real time to provide early warning of security risks. Simultaneously, establish self-optimization performance evaluation metrics such as accuracy and mean squared error. Regularly evaluate the effectiveness of self-optimization and its impact on security, so that problems can be identified and countermeasures implemented promptly.

[0073] S41: Real-time collection of security data and algorithm operation status is used to quantitatively assess data security risks and generate a dynamic risk score. Based on the collected security data and algorithm operation status, a pre-set risk assessment model is used to quantify data security risks. The risk assessment model comprehensively considers multiple factors, each of which is assigned a corresponding weight, and a comprehensive risk score is generated through weighted calculation. The risk of data leakage is assessed based on the sensitivity of the data, the control of access rights, and the effectiveness of network security protection measures.

[0074] S42: Adjust the safety parameters through dynamic risk scoring, calculate the self-optimization parameter compensation value, and obtain the adjusted safety parameters.

[0075] The specific steps for adjusting security parameters through dynamic risk scoring are: S421: Use neural network algorithm to establish a mapping model.

[0076] S422: Obtain the dynamic risk score of the current system and input it into the mapping model. Calculate the value of each security parameter to be adjusted according to the mapping model to obtain the security parameter value to be adjusted.

[0077] S423: Use support vector machine to establish an association model.

[0078] S424: Obtain a safety optimization parameter based on the influence of the value of the safety parameter to be adjusted on the self-optimization parameter, input the safety optimization parameter into the association model to calculate the value of the self-optimization parameter to be adjusted, and obtain the safety self-optimization parameter.

[0079] The calculation formula of the self-optimization parameter compensation value is: ; Where, is the self-optimization parameter after compensation, is the initial self-optimization parameter, is the adjustment coefficient of the kth type of self-optimization parameter, R is the dynamic risk score, is the sensitivity index of risk to parameters.

[0080] S43: Verify the adjusted security parameters in a simulation environment and evaluate their impact on the performance and security of the IoT device algorithm. Construct a simulation environment that resembles an actual IoT system, including IoT devices, network topology, and data transmission protocols. The simulation environment must accurately reflect the actual system's operational conditions to effectively verify the adjusted security parameters. In the simulation environment, virtual devices and simulation software can be used to simulate the behavior of IoT devices. A network simulator can be used to simulate various network conditions, such as bandwidth limitations and latency. Enter the adjusted security parameters into the simulation environment, run the IoT device algorithm, and record the algorithm's performance and security metrics. Performance metrics include algorithm execution time, resource utilization, and output accuracy. Security metrics include the likelihood of data leakage and the detection rate of data tampering. To ensure the reliability of the verification results, conduct multiple experiments and perform statistical analysis on the experimental results. Based on the experimental results, calculate the impact index of the adjusted security parameters on the performance and security of the IoT device algorithm.

[0081] S44: Optimize and improve the safety parameter adjustment plan based on the impact index, re-determine the safety parameter adjustment range, and obtain dynamically adjusted safety parameters.

[0082] S5: Build a security audit system based on dynamically adjusted security parameters to monitor in real time whether the behavior of the IoT device algorithm exceeds the preset threshold, and issue an early warning if it does.

[0083] The security audit system, built based on dynamically adjusted security parameters, integrates multi-source data collection interfaces to aggregate IoT device algorithm behavior logs, communication protocol interaction records, and operational status metadata in real time. Using a stream processing engine (e.g., parsing and feature extraction from raw data streams), it generates real-time behavioral profiles encompassing metrics such as algorithm execution time, resource utilization, and data tampering detection rate. The system also incorporates a decision engine based on a dynamic threshold model. This model correlates the device risk level matrix output by the risk assessment area with the algorithm behavior baseline. It uses a sliding window technique to perform weighted aggregation calculations on behavioral features over the past N cycles. It then integrates an anomaly detection algorithm to identify behavioral patterns that deviate from normal distributions. When the algorithm execution time exceeds 15% of the baseline, the data verification failure rate exceeds the threshold, or the resource utilization rate triggers the dynamic allocation cap, the system automatically triggers an alert process, generating an alert event containing the device ID, algorithm identifier, risk type, timestamp, and associated log fragments. This event is then pushed to the operations and maintenance platform via a message queue. Root cause analysis is then initiated to trace the risk transmission path triggered by the abnormal behavior. Ultimately, the alert response results are linked to the dynamic threshold optimization strategy, forming a closed-loop control system from behavior monitoring to policy iteration, ensuring the secure and compliant operation of IoT algorithms in complex environments.

[0084] In summary, this embodiment provides an algorithm optimization method for IoT devices. By deeply embedding the encryption mechanism into the TFF algorithm framework, combining the differential privacy Laplace mechanism to quantify data, constructing a security-self-optimization parameter association mechanism, dynamically adjusting security parameters, and building a security audit closed-loop system, it solves the core problems faced by IoT algorithms in the process of data transmission, storage, and self-optimization, such as privacy leakage risks, rigid parameter configuration, and delayed security response. On the whole, it ensures the efficient self-optimization of IoT algorithms while accurately assessing data risks, dynamically adapting security parameters, monitoring abnormal behaviors in real time, and issuing timely warnings. Through strategy iteration, a closed-loop control is formed, which greatly enhances the data security of the IoT system and the stability of algorithm operation, effectively responds to complex and changing data security challenges, and helps the reliable development of IoT services.

[0085] Based on the same general inventive concept, the present invention also protects an algorithm optimization system for an Internet of Things device. The algorithm optimization system for an Internet of Things device provided by the present invention is described below. The algorithm optimization system for an Internet of Things device described below and the algorithm optimization method for an Internet of Things device described above can be referenced to each other.

[0086] Figure 2 This is a structural diagram of an algorithm optimization system for an Internet of Things device provided by an embodiment of the present invention.

[0087] like Figure 2 As shown, an algorithm optimization system for IoT devices includes: Algorithm encryption module: embeds the encryption area into the TFF algorithm framework to obtain the encrypted TFF algorithm framework.

[0088] Data quantification module: Based on the encrypted TFF algorithm framework and combined with the Laplace mechanism, the original data is converted into quantifiable data to obtain encrypted quantitative indicators.

[0089] Risk assessment module: Analyzes the mutual influence between security parameters and self-optimization parameters based on encrypted quantitative indicators to obtain data risk assessment.

[0090] Dynamic adjustment module: Dynamically adjusts security parameters during the self-optimization process of the IoT algorithm based on data risk assessment to obtain dynamically adjusted security parameters.

[0091] Synchronous early warning module: Build a security audit system based on dynamically adjusted security parameters. The security audit system analyzes behavioral profiles, synchronizes early warnings, and associates policy iterations for closed-loop control.

[0092] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods of each embodiment or certain portions of the embodiments.

[0093] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will appreciate that modifications may be made to the technical solutions described in the aforementioned embodiments, or that some of the technical features may be replaced with equivalents. Such modifications or replacements do not deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An algorithm optimization method for an Internet of Things device, characterized in that: include: S1: Use the TFF algorithm to build the basic framework of the IoT device algorithm, embed the encrypted area into the TFF algorithm framework, and obtain the encrypted TFF algorithm framework; S2: The input data of the IoT device is raw data. According to the encrypted TFF algorithm framework and combined with the Laplace mechanism, the raw data is converted into encrypted quantitative indicators. S3: Analyze and establish a correlation mechanism between security parameters and self-optimization parameters based on the encryption quantitative indicators to obtain risk assessment data; S4: Dynamically adjust security parameters during the self-optimization process of the IoT algorithm according to the risk assessment data to obtain dynamically adjusted security parameters; S5: A security audit system is built based on the dynamically adjusted security parameters and the behavioral profile is analyzed. When the algorithm execution time exceeds the threshold, the IoT device issues a synchronous warning, thereby obtaining an IoT device algorithm with dynamic self-optimization capabilities and data security.

2. The algorithm optimization method for an Internet of Things device according to claim 1, characterized in that: In step S1, the TFF algorithm framework includes: a layered architecture of the algorithm framework, with the underlying data processing and basic operations as the basic layer; the middle layer is set as the core algorithm layer; and the upper layer is planned as the security and self-optimization management layer.

3. The algorithm optimization method for an Internet of Things device according to claim 1, characterized in that: In step S1, the specific steps of building the TFF algorithm framework for data encryption include: S11: Build the overall architecture of the secure feedback channel based on the characteristics of the TFF algorithm framework for data encryption and business requirements; S12: Integrate an encryption communication protocol in the feedback channel and use a symmetric encryption algorithm to encrypt the transmitted data; S13: Use a combination of digital certificates and biometrics to verify the identity of users and devices.

4. The algorithm optimization method for an Internet of Things device according to claim 1, characterized in that: In step S2, the Laplace mechanism is used to protect privacy information. The specific steps for converting raw data into quantifiable data are as follows: S21: Allocate privacy budget to Laplace mechanism; S22: Adding noise and performing perturbation processing on the original data according to the privacy budget to obtain hidden data; S23: Classifying and standardizing the hidden data; S24: Constructing a mapping relationship between the standardized hidden data and the quantitative indicators; S25: Aggregate and analyze the hidden data according to the mapping relationship to obtain an encrypted quantitative index.

5. The algorithm optimization method for an Internet of Things device according to claim 1, characterized in that: In step S3, the encryption quantitative index analyzes the mutual influence relationship between the security parameters and the self-optimization parameters. The specific steps of establishing the association mechanism are as follows: S31: Extract and analyze the characteristics of the encryption quantitative indicators, sort out the correlation performance of security parameters and self-optimization parameters in various links of data processing, and obtain preliminary correlation data; S32: Construct a logical relationship chain between the safety parameters and the self-optimization parameters to obtain detailed correlation data; S33: Formulate security parameters and self-optimization parameter adjustment rules and strategies corresponding to different impact levels based on the detailed association data and historical data to obtain an association mechanism; S34: Applying the association mechanism to simulate the impact of parameter changes on the overall security status of data under different security scenarios to obtain risk assessment data.

6. The algorithm optimization method for an Internet of Things device according to claim 5, characterized in that: In step S32, the specific steps of building the relationship chain are: S321: Sorting out the complete path of interaction between safety parameters and self-optimization parameters; S322: Draw a logical relationship diagram of the mutual influence between the safety parameters and the self-optimization parameters through visualization; S323: Inputting a quantitative model according to the complete path and the logical relationship diagram, substituting different parameter value ranges, and obtaining detailed correlation data.

7. The algorithm optimization method for an Internet of Things device according to claim 1, characterized in that: In step S4, the specific steps of dynamically adjusting the security parameters are: S41: Real-time collection of security data and algorithm operation status, quantitative assessment of data security risks, and dynamic risk scoring; S42: Adjust the security parameters according to the dynamic risk score, and calculate the self-optimization parameter compensation value to obtain the security self-optimization parameter; S43: Verify the impact index of the security self-optimization parameter on the performance and security of the IoT device algorithm in a simulation environment; S44: Optimizing and improving the security parameter adjustment scheme according to the impact index, re-determining the security parameter adjustment range, and obtaining dynamically adjusted security parameters.

8. The algorithm optimization method for an Internet of Things device according to claim 7, characterized in that: In step S42, the specific steps of adjusting the security parameters using the dynamic risk score are as follows: S421: Use neural network algorithm to establish mapping model; S422: Obtain a dynamic risk score of the current system and input it into the mapping model, and calculate the value to be adjusted for each security parameter according to the mapping model to obtain the security parameter value to be adjusted; S423: Use support vector machine to establish association model; S424: Obtain a safety optimization parameter based on the influence of the safety parameter value to be adjusted on the self-optimization parameter, input the safety optimization parameter into the association model to calculate the value to be adjusted for the self-optimization parameter, and obtain the safety self-optimization parameter.

9. The algorithm optimization method for an Internet of Things device according to claim 1, characterized in that: Building a security audit system includes: S51: Use the stream processing engine to parse and extract features from the original data stream to generate real-time behavior profiles; S52: Based on the risk assessment results and behavioral baseline, a dynamic threshold model is used to determine the real-time monitoring threshold; S53: Use sliding window technology to perform weighted aggregation calculations on behavioral features within the past N cycles, and combine it with anomaly detection algorithms to identify behavioral patterns that deviate from normal distributions. When the algorithm execution time exceeds the threshold, a message is synchronously pushed to the IoT device terminal; S54: Analyze the risk transmission path triggered by abnormal regional tracing behavior, associate the early warning and disposal results with the dynamic threshold optimization strategy, and form a closed-loop management and control system from behavior monitoring to strategy iteration.

10. An algorithm optimization system for an Internet of Things device, which adopts an algorithm optimization method for an Internet of Things device according to any one of claims 1 to 9, characterized in that: The algorithm optimization system includes: Algorithm encryption module: embeds the encrypted area into the TFF algorithm framework to obtain the encrypted TFF algorithm framework; Data quantification module: According to the encrypted TFF algorithm framework, combined with the Laplace mechanism, the raw data is converted into quantifiable data to obtain encrypted quantitative indicators; Risk assessment module: Analyzes the mutual influence relationship between security parameters and self-optimization parameters based on the encryption quantitative indicators to obtain data risk assessment; Dynamic adjustment module: dynamically adjusts security parameters during the IoT algorithm self-optimization process according to the data risk assessment to obtain dynamically adjusted security parameters; Synchronous warning module: A security audit system is built based on the dynamically adjusted security parameters. The security audit system analyzes behavioral profiles, synchronizes warnings, and associates strategy iterations for closed-loop control.

Images