Automatic connection and handshake protocol of trusted data space connector based on AI

Abstract

The invention discloses an AI-based automatic connection and handshake protocol for a trusted data space connector. The AI-based automatic connection and handshake protocol comprises a data supply end connector, a data demand end connector and a platform operation end, the data supply end connector is used for collecting and preprocessing data; the demand end connector is used for receiving and processing data; the platform operation end is used for connection management, authority distribution and security auditing; the method has the beneficial effects that through automatic connection and an intelligent handshake protocol, manual intervention is reduced, and the data circulation cost is reduced; multi-factor authentication, dynamic encryption and privacy protection technologies are adopted to ensure the safety of the whole life cycle of the data; aI and machine learning technologies are introduced, so that the system can intelligently select an optimal strategy according to real-time network conditions and data features; interoperation and data sharing among different data spaces are realized, and circulation and value mining of data elements are promoted.

Description

Technical Field

[0001] The present invention belongs to the technical field of trusted data space, and specifically relates to an AI-based automated connection and handshake protocol for trusted data space connectors. Background Art

[0002] In the digital age, data has become the core element driving social progress and economic development; however, current data circulation faces the following key challenges: ‌Insufficient compatibility of heterogeneous systems‌: Data systems of different manufacturers adopt different transmission protocols, such as HTTP / HTTPS, MQTT, and RTP. Connectors need to manually configure adaptation rules, resulting in low efficiency. ‌Poor adaptability to dynamic networks‌: Fluctuations in 5G / satellite networks lead to a decline in transmission stability. Traditional protocols cannot adjust parameters in real time, and the packet loss rate increases by more than 30%. ‌Conflict between privacy and usability‌: Fixed encryption strategies are difficult to balance data privacy and computing efficiency, and quantum computing threatens the existing encryption system. ‌Lack of standardization‌: The technology of trusted data space has not yet formed a unified standard, and cross-platform interoperability is limited. Existing technologies such as federated learning frameworks and dynamic routing technologies can partially solve the above problems, but lack end-to-end systematic optimization. Summary of the Invention

[0003] The purpose of the present invention is to provide an AI-based automated connection and handshake protocol for trusted data space connectors, solve the problem of compatibility of heterogeneous systems, avoid manually configuring adaptation rules, and achieve seamless docking between different transmission protocol systems; improve the adaptability to dynamic networks, adjust parameters in real time in complex network environments, reduce the packet loss rate, and ensure the stability and continuity of data transmission; balance privacy and usability, cope with the threat of quantum computing, and reduce the risk of data leakage; achieve intelligent decision-making, dynamically adjust connection strategies according to real-time network conditions, data sensitivity, etc.; promote the standardization of trusted data space technology, enhance cross-platform interoperability; improve data life cycle management, ensure the security and traceability of data at each stage; strengthen the ability to handle exceptions, and effectively detect and handle sudden changes in network traffic and abnormal access.

[0004] To achieve the above purpose, the present invention provides the following technical solution: An AI-based automated connection and handshake protocol for trusted data space connectors, including a data-supplying end connector, a data-requiring end connector, and a platform operation end; the data-supplying end connector is used for collecting and preprocessing data; the data-requiring end connector is used for receiving and processing data; the platform operation end is used for connection management, permission allocation, and security auditing.

[0005] As a preferred technical solution of the present invention, it further includes a protocol parsing module, which uses a convolutional neural network to identify protocol header features, combines with a knowledge graph for auxiliary decision-making, and determines the protocol type used.

[0006] As a preferred technical solution of the present invention, it further includes a transmission path selection module, which is based on Bayesian optimization decision-making and reinforcement learning path optimization, dynamically selects a transmission strategy according to network latency and bandwidth utilization, and realizes millisecond-level path switching during network fluctuations.

[0007] As a preferred technical solution of the present invention, it further includes a data adaptation module, which uses data mapping and conversion technology to process heterogeneous data formats, and combines federated learning and differential privacy technology for joint modeling and privacy protection.

[0008] As a preferred technical solution of the present invention, it further includes identity authentication and permission management for the intelligent handshake protocol. The identity authentication and permission management uses multi-factor identity authentication and zk-SNARKs technology to verify permissions, and realizes dynamic permission management through blockchain smart contracts.

[0009] As a preferred technical solution of the present invention, it further includes data encryption and privacy protection for the intelligent handshake protocol. The data encryption and privacy protection adopts a multi-level encryption strategy according to the data sensitivity level, and uses differential privacy technology during data statistical analysis.

[0010] As a preferred technical solution of the present invention, it further includes a handshake negotiation mechanism for the intelligent handshake protocol, which uses the PPO algorithm to optimize the negotiation strategy.

[0011] As a preferred technical solution of the present invention, it further includes a DAG blockchain network and a smart contract; the DAG blockchain network records data hashes, usage events, and destruction proofs, and the smart contract automatically executes data destruction and generates destruction vouchers.

[0012] As a preferred technical solution of the present invention, it further includes an exception handling mechanism, which uses the LSTM algorithm to predict network traffic for dynamic flow limiting, and uses the isolation forest algorithm to detect abnormal access and trigger re-identification.

[0013] Compared with the prior art, the beneficial effects of the present invention are: Through automated connection and intelligent handshake protocol, manual intervention is reduced, and the data circulation cost is lowered; By adopting multi-factor authentication, dynamic encryption, and privacy protection technologies, the security of the entire data life cycle is ensured; By introducing AI and machine learning technologies, the system can intelligently select the optimal strategy according to the real-time network conditions and data characteristics; Implement interoperability and data sharing between different data spaces, and promote the circulation and value mining of data elements. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is the system architecture diagram of the present invention; Figure 2 It is the protocol parsing flowchart of the present invention; Figure 3 It is the handshake negotiation flowchart of the present invention; Figure 4 It is the data processing flowchart of the present invention; Figure 5 It is the exception handling flowchart of the present invention; Figure 6 It is the blockchain evidence storage flowchart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. Embodiment 1

[0016] Please refer to Figures 1-6 , which is the first embodiment of the present invention. This embodiment provides an AI-based trusted data space connector automated connection and handshake protocol, including a data-providing end connector, a data-requiring end connector, and a platform operation end; the data-providing end connector: responsible for collecting data from the data source and performing preprocessing, including protocol adaptation, data cleaning, and encryption; the data-requiring end connector: responsible for receiving and processing data from the data-providing end, including federated learning, homomorphic encryption, and data destruction verification; the platform operation end: as the management and coordination center of the system, responsible for connection management, permission allocation, and security auditing.

[0017] In this embodiment, preferably, it further includes an automated connection mechanism, and this automated connection mechanism includes Protocol Parsing Module: Uses a Convolutional Neural Network (CNN) to learn and identify protocol header features, supports more than 30 protocols (such as HTTP, MQTT, CoAP, OPC UA, etc.), the protocol identification takes less than 1 second, trains a protocol classifier through a CNN model with the ResNet-50 architecture, and the accuracy rate is ≥99.8%; constructs a knowledge graph related to the data space, integrating protocol features, applicable scenarios, data types and their association relationships; when performing CNN protocol parsing and Bayesian optimization decision-making, provides more comprehensive information based on the knowledge graph to improve the accuracy of protocol parsing and transmission strategy selection; Working principle: In the connection establishment stage, the data-supplying end connector uses the CNN protocol parsing module to extract and identify protocol header features to determine the protocol type used; subsequently, the Bayesian optimization decision-making module selects the most suitable transmission strategy according to parameters such as the current network conditions; for example, when the network latency is high, select the QUIC+MQTT protocol to improve transmission stability; when the network conditions are good, select the HTTPS / 2 protocol to improve transmission speed; Transmission Path Selection Module: Based on parameters such as network latency (RTT) and bandwidth utilization (BWU), dynamically selects the optimal transmission strategy (such as enabling MQTT+QUIC for real-time streaming data and using FTP over TLS for batch files), with a 40% improvement in transmission efficiency; uses the Proximal Policy Optimization (PPO) algorithm to construct a network state matrix (including node load, link latency, packet loss rate), and achieves millisecond-level switching of the transmission path in 5G network fluctuation scenarios, reducing the end-to-end latency to less than 200ms; Working principle: During data transmission, the reinforcement learning path optimization module monitors the network state in real time and constructs a network state matrix; when network fluctuations cause the current transmission path to be unstable, the PPO algorithm quickly selects a new optimal transmission path based on the network state matrix to achieve millisecond-level switching and ensure the stability and efficiency of data transmission; Data Adaptation Module: Aims at the data format differences in heterogeneous systems and uses data mapping and conversion technologies to convert data in different formats into a unified standard format; through predefined data mapping rules, can quickly and accurately complete data conversion, improving data compatibility and usability; combines federated learning and differential privacy technologies to perform joint modeling without data leaving the domain; in the model training stage of federated learning, adds noise that conforms to a specific distribution to gradients or model parameters to achieve more refined privacy protection and reduce the risk of data privacy leakage; Working principle: The data-supplying end connector preprocesses and extracts features from the data locally, and then uploads the model parameters (instead of the original data) to the platform operation end for joint modeling; in this way, the model training can be completed without data leaving the domain, protecting data privacy.

[0018] In this embodiment, preferably, it further includes an intelligent handshake protocol, and the intelligent handshake protocol includes Identity Authentication and Permission Management: Multiple factors such as digital certificates and biometric recognition are required for authentication at multiple ends to ensure the authenticity of identities; the zk-SNARKs technology is used to verify the data usage permissions of the demanding parties, with the authentication time consuming less than 50ms and without revealing the original data of the data-providing parties; an ECC temporary session key is dynamically generated, and the key is fragmented and stored in the DAG blockchain in combination with the Shamir secret sharing algorithm, and a single-node attack cannot restore the complete key; the dynamic permission management of data usage is realized by using blockchain smart contracts; during the data usage process, the permissions are adjusted in real time according to preset conditions (such as usage volume, usage time, flow area, etc.), such as restricting access, suspending usage, or requiring re-authentication, to enhance the security and controllability of the data: Working principle: In the dynamic key negotiation stage of connection establishment, the demanding parties initiate zk-SNARKs permission proofs, and the data-providing parties generate ECC temporary keys through federated learning; zero-knowledge proof authentication ensures that the original data of the data-providing parties is not leaked, and the dynamically generated ECC temporary session key is stored in the DAG blockchain in combination with the Shamir secret sharing algorithm to enhance the security of the key and prevent key leakage caused by single-node attacks; Data Encryption and Privacy Protection: Dynamically switch encryption algorithms according to the data sensitivity level to balance privacy and computational overhead; >Level 1: AES-128 + HMAC-SHA256, applicable to public data (meteorology / traffic); >Level 2: AES-256-GCM + Homomorphic Encryption, applicable to medical / financial data; >Level 3: NTRU Quantum-Resistant Algorithm, applicable to national defense / energy confidential data; Differential Privacy Technology: During the data statistics and analysis process, the data is perturbed to protect the privacy of the data; Working principle: Before data transmission, the data-providing party's connector selects the corresponding encryption algorithm according to the data sensitivity level; for public data (such as meteorological and traffic data), AES-128+HMAC-SHA256 encryption is used; for sensitive data such as medical and financial data, AES-256-GCM+homomorphic encryption is used; for confidential data such as national defense and energy, the NTRU quantum-resistant algorithm is used; by dynamically switching encryption algorithms, both the privacy of the data is guaranteed and unnecessary computational overhead is reduced; Handshake Negotiation Mechanism: The PPO algorithm is used to analyze and learn historical handshake data, continuously optimize the handshake negotiation strategy, and improve the handshake efficiency and success rate; Working principle: During the connection establishment process, the data-providing party and the demanding party negotiate through an intelligent handshake protocol to determine parameters such as the data transmission protocol, encryption algorithm, and data format; the PPO algorithm dynamically adjusts the negotiation strategy according to historical handshake data to ensure an efficient and stable handshake process.

[0019] In this embodiment, preferably, it further includes blockchain enhanced evidence storage, and the blockchain enhanced evidence storage includes DAG blockchain network: It uses a directed acyclic graph (DAG) structure to record data hashes (SHA-256), usage events, and proof of destruction. The evidence storage query efficiency is >1000 TPS; Working principle: During the data transmission process, the DAG blockchain network records information such as data hashes, usage events, and proof of destruction in real time; Whenever a data usage or destruction event occurs, the relevant records are written into the DAG blockchain through a smart contract, ensuring that the entire life cycle of the data is traceable and the evidence storage query efficiency is high; Smart contract trigger: Automatically call the DataDestruction() function after data usage is completed, fragmentarily erase the data, and generate a destruction voucher. The erasure verification time <2s; Working principle: When data usage is completed, the smart contract automatically triggers the DataDestruction() function; This function erases the data in a fragmentary manner and generates a destruction voucher; The short erasure verification time ensures the timely destruction and irrecoverability of the data, further enhancing the security of the data.

[0020] In this embodiment, preferably, it further includes an exception handling mechanism, and the exception handling mechanism includes Traffic mutation detection: Use the LSTM algorithm to predict network traffic. When the error >15%, dynamic flow limiting (bandwidth allocation adjustment) is triggered, and the accuracy rate ≥98.5%; During the data transmission process, the LSTM algorithm predicts network traffic in real time; When a network traffic mutation is predicted and the error exceeds 15%, the system automatically triggers the dynamic flow limiting mechanism to adjust the bandwidth allocation to ensure the stability and efficiency of network transmission; Abnormal access interception: Detect abnormal access (when the score > 0.7, trigger identity re-verification), F1-score = 0.92; Working principle: The system uses the isolation forest algorithm to detect access requests; When abnormal access is detected and the score exceeds 0.7, the identity re-verification mechanism is triggered, requiring the visitor's identity to be re-verified to ensure that only legitimate users can access and use the data, improving the security of the system.

[0021] Figure 1 The description is as follows: Data source: The data source of the data provider connector can be a database, a file system, etc.; Data reception: The data receiver connector receives data from the data provider; Connection management: Manage the connection between the data provider and the data receiver, including registration, authentication, and authorization; Permission allocation: Allocate data access permissions according to user roles and permissions; Security Audit: Record the access and usage behaviors of data to ensure the security and compliance of the system; Protocol Adaptation Module: Support the identification and adaptation of multiple protocols; Data Preprocessing Engine: Clean, transform, and encrypt data; Dynamic Encryption Controller: Select encryption algorithms according to the data sensitivity level; Federated Learning Module: Conduct joint modeling without data leaving the domain; Homomorphic Encryption Data Aggregator: Aggregate encrypted data; Destruction Verifier: Ensure timely destruction after data usage; DAG Blockchain Network: Record the hash values, usage events, and destruction proofs of data; Smart Contract Management Module: Manage the creation and execution of smart contracts; Real-time Monitoring Dashboard: Monitor the network status and data transmission in real time.

[0022] Figure 2 The illustration is as follows: Start: The process starts; Data Transmission Request: The data-requiring end initiates a data transmission request; Protocol Identification: Identify the protocol type used for data transmission; CNN Protocol Parsing: Use convolutional neural networks to parse the protocol header features; Network Status Evaluation: Evaluate the current network status, such as latency, bandwidth, etc.; Select Transmission Strategy: Select the optimal transmission strategy according to the network status (such as QUIC+MQTT or HTTPS / 2); Transmission Strategy Execution: Execute the selected transmission strategy; Data Transmission: Transmit data according to the selected strategy; End: The process ends.

[0023] Figure 3 The illustration is as follows: Start: The process starts; The data-requiring end initiates a connection request: The data-requiring end sends a connection request to the platform operation end; The platform operation end verifies the identity: Verify the identity of the data-requiring end; Verification Result: Judge whether the identity verification is successful; Reject Connection: The identity verification fails, and the connection is rejected; The data-supplying end generates an ECC temporary key: The data-supplying end generates a temporary session key; The data-requiring end conducts zk-SNARKs permission proof: The data-requiring end conducts zero-knowledge proof to verify the data usage permission; Platform operation terminal verification permission: Verify the permissions of multiple terminals; Permission verification result: Judge whether the permission verification is successful; Negotiate transmission parameters: Negotiate parameters such as data transmission protocols and encryption algorithms; Intelligent contract records negotiation result: Record the negotiation result in the intelligent contract; Data transmission starts: Start data transmission; End: The process ends.

[0024] Figure 4 The description is as follows: Start: The process starts; Data collection: The data provider connector collects data from the data source; Data preprocessing: Perform preprocessing operations such as cleaning and converting the collected data; Data type: Judge the sensitivity level of the data; Encryption processing: Select the corresponding encryption algorithm according to the data type; Public data: Encrypted with AES-128+HMAC-SHA256; Sensitive data: Encrypted with AES-256-GCM+homomorphic encryption; Confidential data: Encrypted with NTRU quantum-resistant algorithm; Data transmission: Transmit the encrypted data to the data requester; Data requester receives data: The data requester receives the encrypted data; Data decryption: Decrypt the encrypted data; Data usage: Use and analyze the decrypted data; End: The process ends.

[0025] Figure 5 The description is as follows: Start: The process starts; During data transmission: During the data transmission process; Traffic monitoring: Real-time monitor network traffic; Normal: The traffic is normal, continue transmission; Abnormal: The traffic is abnormal, perform LSTM traffic prediction; LSTM traffic prediction: Use the LSTM algorithm to predict network traffic; Prediction error: Judge whether the prediction error exceeds 15%; Dynamic flow limiting: If the error exceeds 15%, trigger dynamic flow limiting; Adjust bandwidth allocation: Adjust the bandwidth allocation to optimize network transmission; Continue transmission: Continue transmission after adjustment; Access Detection: Detect whether the access request is normal; Normal: The access is normal, and the process ends; Abnormal: The access is abnormal, and the Isolation Forest algorithm is used for detection; Isolation Forest Algorithm Detection: Detect abnormal access requests; Abnormal Score Calculation: Calculate the abnormal score; Trigger Identity Re-verification: If the score exceeds 0.7, trigger identity re-verification; Re-verify Identity: Re-verify the identity of the visitor; End: The process ends.

[0026] Figure 6 The illustration is as follows: Start: The process starts; Data Generation: Generate the data to be stored; Data Hash Calculation: Calculate the hash value of the data; Data Usage Event Recording: Record the data usage events; Smart Contract Trigger for Archiving: Trigger the smart contract to write the archived information into the blockchain; Write to DAG Blockchain: Write the archived information into the DAG blockchain; Archived Information Query: Query the archived information to verify the integrity and authenticity of the data; End: The process ends.

[0027] Specific Composition and Working Process of the CNN Protocol Analysis Module: Detailed Structure of the Convolutional Neural Network Model: The CNN protocol analysis module adopted in the present invention is based on the ResNet-50 architecture, which contains multiple residual blocks. Each residual block consists of a convolutional layer, a batch normalization layer, and a ReLU activation function. The input of the network is the protocol header data. After feature extraction by the convolutional layer, the final output is the classification result of the protocol type, supporting the identification of more than 30 protocols (such as HTTP, MQTT, CoAP, OPC UA, etc.). The protocol identification takes less than 1s. The protocol classifier is trained through the CNN model of the ResNet-50 architecture, and the accuracy rate is ≥ 99.8%.

[0028] Model Training Method and Data: The training dataset is sourced from the header data of different protocols in the actual network. After preprocessing and annotation, it is used for model training. During the training process, the cross-entropy loss function is used as the loss function, the stochastic gradient descent algorithm is used as the optimization algorithm, the learning rate is set to 0.001, and the training iteration times are 100 epochs.

[0029] Specific processing method for protocol header features: When extracting and identifying protocol header features, first perform normalization processing on the protocol header data to eliminate the influence of data dimensions. Then, extract features through convolutional layers and pooling layers. The convolutional layer uses a 3×3 convolutional kernel, with a stride of 1 and a padding method of same. After the extracted features are activated by the ReLU activation function, they are sent to the fully connected layer for classification and judgment.

[0030] Specific composition and working process of the reinforcement learning path optimization module: Detailed structure of the reinforcement learning algorithm: Adopt a reinforcement learning algorithm based on the Deep Q-Network (DQN). The agent is the core component of this algorithm; there is an interaction relationship between the agent and the environment. The agent selects an action (i.e., the transmission path) according to the current network state, and the environment returns a reward signal and a new network state based on this action. The state space includes parameters such as network latency (RTT) and bandwidth utilization (BWU). The action space includes different transmission path selections. The reward function is designed as follows: When the selected transmission path can minimize the data transmission latency and the packet loss rate, a positive reward is given to the agent; otherwise, a negative reward is given.

[0031] Training process of the reinforcement learning algorithm: The training data includes network state data and transmission path data. The training objective is to enable the agent to select the optimal transmission path under different network states; adopt the experience replay technology to store and replay the training data to improve the training efficiency of the model; the optimization algorithm uses the Adam optimizer, with the learning rate set to 0.0001 and the number of training iterations being 500 episodes.

[0032] Specific construction method of the network state matrix: Construct the network state matrix according to parameters such as network latency (RTT) and bandwidth utilization (BWU). The dimension of the matrix is n×m, where n is the number of network nodes and m is the number of network links. The meaning of the matrix elements is the latency and bandwidth utilization of the network link. Relevant parameters are obtained in real time through network monitoring tools and mapped to specific positions in the matrix.

[0033] Embodiment 2

[0034] Please refer to Figures 1-6 , which is the second embodiment of the present invention. This embodiment is based on the previous embodiment. The difference is: ① Implementation of the data supply end connector Data collection Collection Modes: Support real-time collection, scheduled collection, and manual collection, and adapt to multiple data sources (such as databases and file systems); Hardware Requirements: Deployed on edge nodes, configured with a 4-core CPU, 16GB of memory, and a 100Gbps network interface, running the Ubuntu 22.04 LTS system; Software Dependencies: pip install tensorflow==2.12.0 pyzmq==23.2.1 Data Preprocessing Anomaly Detection and Repair: Use the Isolation Forest algorithm to repair abnormal data and remove noise and redundant information; Data Encryption: Select the corresponding encryption algorithm according to the data sensitivity level and encrypt the data; Working Principle: After data collection, the data provider connector cleans and transforms the data to remove noise and redundant information; then selects an encryption algorithm (such as AES-128, AES-256-GCM, or NTRU) according to the data sensitivity level to encrypt the data; Connection Establishment Protocol Identification: Extract and identify protocol header features through the CNN protocol parsing module to determine the protocol type used; Transmission Strategy Selection: The Bayesian optimization decision-making module selects the optimal transmission strategy according to the network conditions and data characteristics; Working Principle: The data provider connector sends a connection request to the platform operation end, and the platform operation end verifies and authorizes the identity of the data provider; after successful verification, the data provider connector establishes a connection with the data requester connector and performs handshake negotiation; ② Implementation of the Data Requester Connector Data Reception Data Decryption and Verification: Decrypt and verify the received data to ensure data integrity and security; Working Principle: The data requester connector receives data from the data provider, decrypts the data using the corresponding decryption algorithm, and verifies the data integrity; Data Processing Federated Learning Joint Modeling: Combine federated learning technology for joint modeling to generate analysis reports and decision-making suggestions; Working Principle: The data requester connector processes and analyzes the data locally to generate model parameters; then uploads the model parameters to the platform operation end for joint modeling to generate a global model; Connection Management Network Condition Monitoring: Real-time monitor the network condition, dynamically adjust the transmission strategy, and handle abnormal situations; Working principle: During the connection process of the data-requiring end connectors, the network conditions (such as RTT, BWU, packet loss rate) are monitored in real time, and the transmission strategy is dynamically adjusted according to the network conditions to ensure the stability and efficiency of data transmission; ③ Implementation of the platform operation end Connection management Connection registration and authentication: Record all connection information, and manage the registration, authentication, and authorization of the data-providing end and the data-requiring end; Working principle: The platform operation end receives the connection requests from the data-providing end and the data-requiring end, verifies and authorizes the identities of both parties; after the verification passes, record the connection information for convenient query and statistics; Permission allocation Role-based access control: Allocate and manage the access and use of data according to the roles and permissions of users.

[0035] Working principle: The platform operation end allocates corresponding data access permissions according to the roles and permissions of users; only users with corresponding permissions can access and operate sensitive data; Security audit Data operation records: Monitor and audit the access and use of data in real time, and record all data operation behaviors; PBFT consensus algorithm: Adopt the Practical Byzantine Fault Tolerance (PBFT) consensus algorithm to ensure the integrity and immutability of audit records; Working principle: The platform operation end monitors the access and use behaviors of data in real time, and records all operation behaviors (such as query, modification, deletion); these records are written into the DAG blockchain through the PBFT consensus algorithm to ensure the authenticity and immutability of audit records; generate security audit reports regularly to provide a basis for the security management of the system.

[0036] Although the embodiments of the present invention have been shown and described, see the above detailed description. For those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An AI-based automated connection and handshake protocol for a trusted data space connector, characterized in that: It includes a data provider connector, a data requester connector, and a platform operation terminal; the data provider connector is used for collecting and preprocessing data; the data requester connector is used for receiving and processing data; the platform operation terminal is used for connection management, permission allocation, and security auditing.

2. The automated connection and handshake protocol of a trusted data space connector based on AI according to claim 1, characterized in that: It also includes a protocol parsing module, which uses a convolutional neural network to identify protocol header features, combines a knowledge graph for auxiliary decision-making, and determines the protocol type in use.

3. An AI-based trusted data space connector automated connection and handshake protocol according to claim 1, characterized in that: It also includes a transmission path selection module, which is based on Bayesian optimization decision-making and reinforcement learning path optimization, dynamically selects a transmission strategy according to network latency and bandwidth utilization, and realizes millisecond-level path switching during network fluctuations.

4. An AI-based automated connection and handshake protocol for a trusted data space connector according to claim 1, characterized in that: It also includes a data adaptation module, which uses data mapping and conversion technologies to process heterogeneous data formats, and combines federated learning and differential privacy technologies for joint modeling and privacy protection.

5. The automated connection and handshake protocol of a trusted data space connector based on AI according to claim 1, characterized in that: It also includes identity authentication and permission management for the intelligent handshake protocol. Identity authentication and permission management use multi-factor identity authentication and zk-SNARKs technology to verify permissions, and realize dynamic permission management through blockchain smart contracts.

6. An AI-based automated connection and handshake protocol for a trusted data space connector according to claim 1, characterized in that: It also includes data encryption and privacy protection for the intelligent handshake protocol. Data encryption and privacy protection adopt a multi-level encryption strategy according to the data sensitivity level, and use differential privacy technology during data statistical analysis.

7. An AI-based automated connection and handshake protocol for a trusted data space connector according to claim 1, characterized in that: It also includes a handshake negotiation mechanism for the intelligent handshake protocol, which uses the PPO algorithm to optimize the negotiation strategy.

8. An AI-based automated connection and handshake protocol for a trusted data space connector according to claim 1, characterized in that: It also includes a DAG blockchain network and smart contracts; the DAG blockchain network records data hashes, usage events, and destruction proofs, and the smart contracts automatically execute data destruction and generate destruction vouchers.

9. The automated connection and handshake protocol of a trusted data space connector based on AI according to claim 1, characterized in that: It also includes an exception handling mechanism, which uses the LSTM algorithm to predict network traffic for dynamic flow control, and uses the isolation forest algorithm to detect abnormal access and trigger identity re-verification.

Images