An identity authentication method and system based on a national secret algorithm

Abstract

The application discloses an identity authentication method and system based on a national secret algorithm, and comprises the following steps: a layered key management system based on a national secret IBE framework is established, a system master key pair is generated by using an SM2 algorithm, and a decentralized key distribution mechanism is realized; a domain perception differential privacy protection module is constructed, a privacy budget allocation strategy is defined according to a security level, and calibrated Laplace noise is added to a user identity feature vector; a distributed batch matrix multiplication protocol is designed, the secret sharing technology is used to decompose the protocol to multiple computing nodes for parallel processing; a zero-knowledge proof verification mechanism is implemented, and identity authentication is completed through a commitment scheme based on an SM3 hash algorithm; an adaptive key update strategy is deployed, a threat level change is analyzed through a threat situation assessment function; a secure communication channel based on an SM4 symmetric encryption is established, and temporary session keys are used for encryption processing. The application improves the security and scalability of the system, effectively protects the privacy of users, and adapts to a dynamically changing network threat environment.

Description

Technical Field

[0001] This invention relates to the field of network security technology, and in particular to an identity authentication method and system based on national cryptographic algorithms, which is applied to the identity authentication and access control systems of critical information infrastructures such as government affairs, finance, and power. Background Technology

[0002] With the deepening of digital transformation, identity authentication and key management have become core infrastructures of cybersecurity. Under the guidance of the national information security strategy, identity authentication systems based on national cryptographic algorithms have been widely used in critical information infrastructures such as government affairs, finance, and power, which has placed higher demands on the secure distribution and regular updating of keys.

[0003] Existing key management technologies mainly include PKI-based certificate management systems and centralized management schemes based on symmetric keys. PKI systems achieve identity authentication through digital certificates, but suffer from problems such as complex certificate management and difficulty in revocation; centralized key management schemes, while facilitating unified control, face single-point-of-failure risks and scalability limitations. In recent years, Identity-Based Encryption (IBE) technology has provided a new solution for key management.

[0004] The core principle of traditional IBE schemes is to use user identity information as the public key and generate a corresponding private key for the user through a key generation center (PKG). This scheme eliminates the need for digital certificates and simplifies the key distribution process. However, existing technologies still have significant shortcomings in practical applications: first, they lack privacy protection mechanisms in cross-domain identity authentication scenarios, which can easily lead to the leakage of sensitive user information; second, the key update strategy is relatively static and cannot be dynamically adjusted according to the actual threat situation, affecting the security and efficiency of the system. Therefore, developing an identity authentication system that integrates national cryptographic algorithms and supports privacy protection and dynamic key management has significant technical value and application significance. Summary of the Invention

[0005] The purpose of this invention is to provide an identity authentication method and system based on Chinese national cryptographic algorithms. By establishing a hierarchical key management system based on the Chinese national cryptographic IBE framework, and combining differential privacy protection mechanism, distributed batch key update protocol, zero-knowledge proof verification and adaptive key update strategy, the invention solves the problems of insufficient privacy protection and static key update strategy in the prior art.

[0006] To achieve the above objectives, this invention provides an identity authentication method based on Chinese national cryptographic algorithms, comprising: establishing a hierarchical key management system based on the Chinese national cryptographic IBE framework, obtaining elliptic curve parameters and base point G, generating system master key pairs using the SM2 algorithm and establishing domain key generation centers for each identity domain to realize a decentralized key distribution mechanism; based on the decentralized key distribution mechanism, constructing a domain-aware differential privacy protection module, obtaining security level information of user identity domains, defining a privacy budget allocation strategy according to the security level information, and adding calibrated Laplace noise to the user identity feature vector to generate privacy-protected identity authentication data; based on the privacy-protected identity authentication data, designing a distributed batch matrix multiplication protocol to obtain key information for multiple users. New requests are broken down into multiple computing nodes for parallel processing using secret sharing technology to generate batch-updated user keys. Based on these batch-updated user keys, a zero-knowledge proof verification mechanism is implemented. A zero-knowledge proof protocol is constructed using a commitment scheme based on the SM3 hash algorithm to complete identity verification and generate authentication results without exposing the user's private key. Based on the authentication results, an adaptive key update strategy is deployed. Network threat situation data is acquired, and threat level changes are analyzed using a threat situation assessment function to obtain dynamically adjusted key update parameters. Based on these dynamically adjusted key update parameters, a secure communication channel based on SM4 symmetric encryption is established, and temporary session keys are used for encryption to ensure data transmission security during key distribution and update.

[0007] Preferably, the establishment of a hierarchical key management system based on the national cryptographic IBE framework includes:

[0008] Based on the SM2 elliptic curve algorithm and preset curve parameters, the system master key pair is generated through the root key generation center, and a secure elliptic curve base point is selected to obtain the system-level master private key and master public key.

[0009] Based on the system-level master private key and the domain identifiers of each identity domain, a dedicated domain master key is calculated for each identity domain through key derivation functions and SM3 hash operations, and multiple domain key generation centers are established.

[0010] Based on the user's identity identifier and the domain master key, the user's identity is mapped to a point on an elliptic curve through a double hash structure and scalar operations are performed in conjunction with the domain master key to generate a user-specific private key and realize the decentralized key distribution mechanism.

[0011] Preferably, the step of calculating a dedicated domain master key for each identity domain using a key derivation function and SM3 hash operation includes:

[0012] Based on the system-level master private key and the domain identifiers of each identity domain, the domain identifiers are hashed using a hash function based on SM3 to obtain the domain feature value, thereby generating a unique identifier parameter for the domain.

[0013] Perform elliptic curve operations on the domain feature value and the system master private key to generate a domain master private key specific to each identity domain;

[0014] Based on the domain master private key and elliptic curve basis points, the corresponding domain master public key is calculated through elliptic curve dot product operations, thereby establishing a domain-level key management system that supports the decentralized key distribution mechanism.

[0015] Preferably, the construction of the domain-aware differential privacy protection module, which acquires the security level information of the user identity domain, defines a privacy budget allocation strategy based on the security level information, and adds calibrated Laplacian noise to the user identity feature vector to generate privacy-protected identity authentication data, includes:

[0016] Based on the decentralized key distribution mechanism and the data sensitivity of each identity domain, the privacy requirements of information within the domain are analyzed through a security level assessment algorithm to determine a tiered privacy budget allocation strategy.

[0017] Based on the privacy budget allocation strategy and user identity feature vector, identity feature data with added privacy protection noise is obtained by calculating function sensitivity and generating calibrated Laplace noise;

[0018] Based on the identity feature data with added privacy-preserving noise, a heterogeneous joint learning framework is used to collaboratively train an identity recognition model across different domains. Gradient compression and a secure aggregation protocol are employed to protect the model parameter exchange process, thereby obtaining the privacy-preserving identity feature data.

[0019] Preferably, the step of calculating the function sensitivity and generating calibrated Laplace noise includes:

[0020] Based on the privacy budget allocation strategy and the various dimensions of the user identity feature vector, the function sensitivity values ​​of various identity features are determined through theoretical analysis and experimental verification, and the noise scale calculation parameters are obtained.

[0021] Based on the noise scale calculation parameters and the privacy budget parameters in the privacy budget allocation strategy, calibrated Laplace noise is generated according to the ratio of the noise scale calculation parameters to the privacy budget parameters.

[0022] The calibrated Laplace noise is added to each dimension of the user identity feature vector by vector addition, resulting in the identity feature data with added privacy protection noise.

[0023] Preferably, the design of the distributed batch matrix multiplication protocol, which obtains key update requests from multiple users and decomposes them into multiple computing nodes for parallel processing through secret sharing technology to generate batch-updated user keys, includes:

[0024] Based on the privacy-protected identity authentication data and key update requests from multiple users, serial key updates are transformed into batch matrix operations by constructing a user identity matrix and a key generation matrix, resulting in a matrix-based key generation task.

[0025] Based on the matrixed key generation task, the user identity matrix and the key generation matrix are decomposed into multiple matrix shards and distributed to different computing nodes through the Shamir secret sharing scheme to achieve distributed storage of sensitive information.

[0026] Based on matrix sharding of each computing node, the multiplication results of each shard are calculated by performing local matrix multiplication operations in parallel, and the final result is reconstructed using Lagrange interpolation to generate the batch-updated user key.

[0027] Preferably, the implementation of the zero-knowledge proof verification mechanism involves constructing a zero-knowledge proof protocol using a commitment scheme based on the SM3 hash algorithm, to complete identity verification and generate authentication results without exposing the user's private key, including:

[0028] Based on the batch-updated user keys and preset elliptic curve parameters, the corresponding identity public key is generated through discrete logarithm operation and stored in the authentication server, establishing a public-private key correspondence.

[0029] Based on the randomly selected temporary value, the elliptic curve parameters, and the user's private key in the public-private key correspondence, a response value containing the user's private key information is generated by calculating the commitment value and receiving a random challenge from the verifier.

[0030] Based on the commitment value, the random challenge, and the response value, the correctness of the zero-knowledge proof equation is verified using the SM3 hash algorithm, and the authentication result is generated without exposing the user's private key.

[0031] Preferably, the deployment of the adaptive key update strategy, acquiring network threat situation data, analyzing threat level changes through a threat situation assessment function, and obtaining dynamically adjusted key update parameters includes:

[0032] Based on the authentication results and the number of abnormal network logins and attack frequency, a quantified threat level score is calculated using a threat situation assessment function to generate a multi-dimensional threat feature vector.

[0033] Based on the multidimensional threat feature vector and historical threat data, an adaptive grid sampling strategy is used to perform dense sampling in areas where the threat level score exceeds a preset threshold, and a change point detection algorithm is used to identify key node information of threat situation changes, thereby determining the triggering time for key updates.

[0034] Based on the threat level score and the key node information, the threat development trend is predicted through a multi-level threat threshold system and a local function approximation method, and the key update frequency and strength parameters are dynamically adjusted.

[0035] Preferably, establishing a secure communication channel based on SM4 symmetric encryption includes:

[0036] Based on the dynamically adjusted key update parameters and the IBE mechanism, a temporary symmetric encryption key is established by generating a random session key and securely distributing it to both communicating parties using IBE encryption technology.

[0037] Based on the temporary symmetric encryption key and the key distribution data to be transmitted, the data is encrypted using the SM4 symmetric encryption algorithm to obtain the encrypted data.

[0038] Based on the encrypted data and the SM4-GCM authentication encryption mode, the integrity of the data is verified by calculating the message authentication code and attaching it to the encrypted data, thereby preventing tampering attacks during transmission and ensuring the security of data transmission during key distribution and update.

[0039] Preferably, it also includes a blockchain-based decentralized key escrow solution, comprising:

[0040] Based on the decentralized key distribution mechanism, a consortium blockchain network is constructed and the key generation and distribution process is recorded on the blockchain, generating a key management blockchain record containing timestamps and digital signatures to ensure the transparency and immutability of the key management process;

[0041] For the blockchain key management records, a smart contract is designed to automate the key lifecycle management, including periodic updates, permission changes, and emergency revocation operations, and to generate smart contract execution logs and state change records;

[0042] Based on the smart contract execution log, the state change record, and the multi-signature mechanism, multiple authorized nodes are required to participate in key key operations. The blockchain consensus mechanism is used to verify the legality of key operations, establish a distributed trust system, and enhance the security of the decentralized key distribution mechanism.

[0043] Preferably, it also includes a quantum-safe post-quantum cryptographic enhancement scheme, comprising:

[0044] Based on the batch-updated user key and the NTRU lattice base algorithm, a quantum-resistant hybrid key system is constructed by integrating lattice-based cryptographic algorithms and encoding-based cryptographic algorithms, generating quantum-safe user private keys and a quantum threat assessment model;

[0045] Based on the availability of quantum channels, the requirements of traditional key management, and the quantum-safe user private key, a quantum-safe key transmission mechanism is established by designing a fusion protocol of quantum key distribution and traditional key management, which prioritizes the use of quantum key distribution when quantum channels are available.

[0046] Based on the quantum threat assessment model, the quantum-secure key transmission mechanism, and the current cryptographic algorithm strength, the quantum security of the batch-updated user keys is improved by dynamically adjusting the key length and encryption parameters of the quantum cryptographic algorithm.

[0047] Preferably, it also includes a federated learning-based intelligent threat detection enhancement mechanism, including:

[0048] Based on the threat situation analysis results of the dynamically adjusted key update parameters, distributed threat detection nodes are deployed and threat identification models are trained collaboratively using federated learning technology to generate threat level assessment results and threat classification labels.

[0049] Based on the threat level assessment results and the threat classification labels, a multi-level threat response mechanism is designed to automatically adjust the key update policy, access control policy and audit policy according to the threat level, and generate threat response execution records and system security configuration adjustment logs.

[0050] Based on the threat response execution records, the system security configuration adjustment logs, historical attack patterns, and the current system status, predictive security analysis is performed and an artificial intelligence decision engine is integrated to predict potential security risks and deploy protective measures in advance.

[0051] This invention also provides an identity authentication system based on national cryptographic algorithms, comprising:

[0052] The hierarchical key management module is used to establish a hierarchical key management system based on the national cryptographic IBE framework, obtain elliptic curve parameters and base point G, use the SM2 algorithm to generate system master key pairs and establish domain key generation centers for each identity domain, thereby realizing a decentralized key distribution mechanism.

[0053] The privacy protection module is used to construct a domain-aware differential privacy protection module based on the decentralized key distribution mechanism, obtain the security level information of the user identity domain, define a privacy budget allocation strategy according to the security level information, add calibrated Laplace noise to the user identity feature vector, and generate privacy-protected identity authentication data.

[0054] The batch key update module is used to design a distributed batch matrix multiplication protocol based on the privacy-protected identity authentication data, obtain key update requests from multiple users, decompose them to multiple computing nodes for parallel processing through secret sharing technology, and generate batch updated user keys.

[0055] The zero-knowledge verification module is used to implement a zero-knowledge proof verification mechanism based on the batch-updated user keys. It constructs a zero-knowledge proof protocol through a commitment scheme based on the SM3 hash algorithm, and completes identity verification and generates authentication results without exposing the user's private key.

[0056] The adaptive policy control module is used to deploy an adaptive key update policy based on the authentication result, acquire network threat situation data, analyze the changes in threat level through a threat situation assessment function, and obtain dynamically adjusted key update parameters.

[0057] The secure communication module is used to establish a secure communication channel based on SM4 symmetric encryption according to the dynamically adjusted key update parameters, and to use a temporary session key for encryption to ensure the security of data transmission during key distribution and update.

[0058] The beneficial effects of this invention are as follows: by integrating national cryptographic algorithms with modern cryptographic technology, a highly secure and efficient identity authentication system is constructed; through differential privacy mechanism, the privacy and security of user identity information during cross-domain authentication are effectively protected; through distributed batch processing technology, the efficiency of key updates is significantly improved; and through adaptive key update strategy, intelligent response to network threats is achieved, thereby improving system security and resource utilization efficiency. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0060] Figure 1 This is a flowchart illustrating the identity authentication method based on the national cryptographic algorithm of the present invention;

[0061] Figure 2 This is a schematic diagram of the identity authentication system based on the national cryptographic algorithm of the present invention. Detailed Implementation

[0062] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0063] Example 1

[0064] like Figure 1 As shown, an identity authentication method based on Chinese cryptographic algorithms includes the following steps:

[0065] Step S1: Establish a hierarchical key management system based on the national cryptographic IBE framework, obtain elliptic curve parameters and base point G, use the SM2 algorithm to generate system master key pairs and establish domain key generation centers for each identity domain to realize a decentralized key distribution mechanism.

[0066] The IBE framework is an identity-based encryption system based on domestic commercial cryptographic algorithm standards, allowing the use of user identity information (such as email addresses and mobile phone numbers) as public keys. This framework integrates domestic cryptographic algorithms (SM2 / SM3 / SM4) with IBE technology, providing encryption solutions that meet domestic security requirements. The hierarchical key management system allocates key management authority to different levels of entities, including system-level (holding the root master key), domain-level (managing identity domains), and user-level, reducing the risk of single points of failure and improving system scalability.

[0067] Elliptic curve parameters are a set of mathematical parameters that define the security characteristics of a cryptographic system, including the prime number p, equation parameters a / b, the base point G, and its order n. In the Chinese national cryptographic standard, the SM2 curve uses 256-bit prime field parameters, providing security strength comparable to RSA 3072 bits. The base point G is a specific point on the elliptic curve; all cryptographic operations are based on this point and are fundamental to key generation. Specifically, the base point G is a public parameter in elliptic curve cryptography, referring to a large prime-order generator point of order n on the selected elliptic curve. In the Chinese national cryptographic standard SM2 algorithm, the coordinates of the base point G (x, y, y) are... G , y G ) is a predefined, fixed value that conforms to a curve equation and possesses good cryptographic properties. All cryptographic operations, including public key generation (via scalar multiplication ms), are performed on this public key. k The point operations in the key generation and cryptographic operation system all take the base point G as the starting point and benchmark, and therefore it is the foundation of the entire key generation and cryptographic operation system.

[0068] The SM2 algorithm is an elliptic curve public-key algorithm that provides digital signatures, key exchange, and encryption. An identity domain is a group of users with shared management needs, such as enterprise departments or related organizations. The domain key generation center is responsible for verifying the identities of users within the domain and generating corresponding private keys, performing these operations using a domain master key derived from the system master key. The decentralized key distribution mechanism is implemented through multiple independent but collaborative domain centers, reducing the risk of single points of failure, improving system scalability, and allowing users to obtain services from the nearest domain center, thus enhancing efficiency and privacy protection.

[0069] Step S2: Based on the decentralized key distribution mechanism, construct a domain-aware differential privacy protection module, obtain the security level information of the user identity domain, define a privacy budget allocation strategy according to the security level information, add calibrated Laplace noise to the user identity feature vector, and generate privacy-protected identity authentication data.

[0070] The domain-aware differential privacy protection module is a mechanism that adaptively adjusts the strength of privacy protection based on the security requirements of different identity domains. Differential privacy is a technique that protects personal data by adding calibrated random noise, ensuring that the analysis results do not significantly depend on the data of any single user. Building upon standard differential privacy, domain-aware differential privacy introduces the ability to perceive the characteristics of the user's domain, setting different protection parameters for domains with different security levels.

[0071] Security level information is a quantitative description of the security requirements for an identity domain, typically categorized into multiple levels (low, medium, high, very high, etc.), determined based on data sensitivity, regulatory requirements, and risk assessment. The privacy budget allocation strategy determines how differential privacy parameters (ε) are assigned to different domains and data types. Smaller ε values ​​provide stronger privacy protection but may reduce data utility; high-security-level domains are usually assigned smaller ε values ​​to provide stronger protection.

[0072] A user identity feature vector is a multidimensional data structure describing user identity attributes, which may include biometrics, behavioral patterns, or knowledge factors. Laplace noise is a commonly used type of random noise in differential privacy. The addition steps include: calculating feature sensitivity, determining noise parameters based on the privacy budget, generating and adding noise to the original features. The calibration process ensures that the noise level is appropriate, providing sufficient privacy protection without unduly affecting authentication accuracy. Privacy-preserving identity authentication data retains enough information for accurate authentication while preventing unauthorized parties from extracting precise user identity features.

[0073] Step S3: Based on the privacy-protected identity authentication data, design a distributed batch matrix multiplication protocol, obtain key update requests from multiple users, decompose them into multiple computing nodes for parallel processing through secret sharing technology, and generate batch updated user keys;

[0074] Distributed batch matrix multiplication protocols allow multiple computing nodes to securely collaborate on large-scale matrix operations without any single node possessing the complete input data. In identity authentication systems, key update operations are represented as matrix multiplication, enabling batch processing of multiple user requests and significantly improving efficiency. A key update request is a key replacement application triggered by key rotation policies, security events, or permission changes, and includes authentication information and the current key identifier.

[0075] Secret sharing technology breaks down sensitive data into multiple shares distributed to different participants. Only when a sufficient number of shares are combined can the original data be reconstructed. The commonly used Shamir threshold scheme is based on polynomial interpolation to implement an (t,n) threshold mechanism, meaning that any t shares out of n can reconstruct the secret. In distributed computing, secret sharing ensures that no single node can obtain the complete information, protecting data security.

[0076] Computation nodes are servers that execute processing tasks. In this method, multiple nodes hold a portion of the data and collaborate through secure communication to complete the computation. Parallel processing achieves high efficiency through task decomposition, data distribution, synchronous computation, and result reconstruction, making it particularly suitable for large-scale batch updates of user keys. Batch updates of user keys involve updating multiple user keys simultaneously in a single operation, which is more efficient than individual processing while maintaining security and consistency.

[0077] Step S4: Based on the batch-updated user keys, implement a zero-knowledge proof verification mechanism. Construct a zero-knowledge proof protocol using a commitment scheme based on the SM3 hash algorithm to complete identity verification and generate authentication results without exposing the user's private key.

[0078] Zero-knowledge proof verification mechanisms allow users to prove they possess the correct private key without actually revealing its contents. Its core characteristics include: completeness (a true statement convinces the verifier), reliability (a false statement is virtually impossible to succeed), and zero-knowledge (the verifier's information is limited to the statement's truthfulness). In identity authentication, users prove they possess the correct private key without exposing the key itself, preventing credential leakage.

[0079] The SM3 hash algorithm is a national cryptographic standard hash algorithm that outputs a 256-bit hash value. It is similar in design to SHA-256 but with additional security enhancements. In zero-knowledge proofs, it is used to construct commitment schemes, converting secret values ​​into fixed-length commitment values. A commitment scheme consists of two phases: a commitment phase (the user chooses a secret value m and a random number r, and calculates C=SM3(m||r)) and a disclosure phase (the user publicly reveals m and r for verification). It possesses binding properties (the committed value cannot be changed) and concealment properties (the committed value does not reveal the original value).

[0080] The zero-knowledge proof protocol implementation steps include: initializing system parameters; the user generating a commitment value and sending it to the verifier; the verifier sending a random challenge; the user calculating the response; and the verifier checking the validity of the response. Through multiple rounds of interaction, the verifier is convinced that the user possesses the correct private key without obtaining the private key information. The authentication result includes the verification status and session metadata, which is signed to ensure integrity and used for subsequent access control and auditing.

[0081] Step S5: Based on the authentication result, deploy an adaptive key update strategy, obtain network threat situation data, analyze the changes in threat level through a threat situation assessment function, and obtain dynamically adjusted key update parameters;

[0082] The adaptive key update policy dynamically adjusts key update behavior based on environmental changes, encompassing a policy engine, monitoring module, decision-making module, and execution module. Compared to fixed-period updates, the adaptive policy responds to real-time security needs, improving system security and resource efficiency. Network threat posture data is comprehensive information describing the current security environment, including attack attempts, vulnerability information, anomalous activity, and intelligence reports, sourced from intrusion detection systems, firewall logs, and threat intelligence platforms.

[0083] The threat posture assessment function converts threat data into quantitative indicators, considering threat severity, prevalence, system exposure, and asset value, and outputs a standardized threat level (such as a 1-10 scale or low / medium / high classification). Changes in threat levels are identified through trend analysis, threshold detection, and rate of change analysis, distinguishing between emerging threats requiring immediate response and persistent threats requiring strategic adjustments.

[0084] The dynamically adjusted key update parameters include configuration values ​​for update frequency, key strength, update range, and verification strength. The adjustment logic follows these principles: increase the update frequency when the threat level rises; prioritize updating relevant components when specific attacks increase. Specific attacks refer to attack patterns that the system can identify and classify through the security monitoring module. These specific types include, but are not limited to: distributed denial-of-service attacks, man-in-the-middle attacks, replay attacks, cryptographic side-channel attacks (such as timing attacks or power consumption analysis targeting the private key calculation process), targeted key guessing, or brute-force attempts. These attacks have detectable characteristic patterns (such as abnormal traffic characteristics, repeated authentication request patterns, suspicious request timing, etc.), and will trigger corresponding security response mechanisms based on their type, strength, and target. For example, these mechanisms may prioritize updating key components affected by the attack, strengthen the protection level of relevant nodes, or temporarily isolate suspicious connections; balancing security requirements with system performance impact. This dynamic adjustment ensures that the system responds appropriately to changing threat environments and maintains the best security posture.

[0085] Step S6: Based on the dynamically adjusted key update parameters, establish a secure communication channel based on SM4 symmetric encryption, and use a temporary session key for encryption to ensure data transmission security during key distribution and update.

[0086] The SM4 symmetric encryption algorithm is a block cipher in the national cryptographic standard, using a 128-bit key and 128-bit blocks, and employing a 32-round nonlinear iterative structure. SM4 boasts advantages such as high efficiency, strong security, and domestic production capabilities, and is used in secure communication to encrypt data to ensure confidentiality and integrity. The steps for establishing a secure communication channel include: authentication, key negotiation, channel initialization, encrypted communication, and session maintenance. Core security attributes include data confidentiality, integrity, authentication, and non-repudiation.

[0087] Temporary session keys are short-term keys used exclusively for a single communication and expire after the session ends. The advantages of using temporary keys include: limiting the scope of potential leaks, providing forward security, and simplifying key management. Session keys are typically generated based on random numbers from both communicating parties, long-term key derivations, and session identifiers, protecting the transmission of sensitive key materials during key distribution.

[0088] The encryption process includes: data segmentation, key expansion, initialization vector generation, encryption operations, runtime mode processing, and integrity protection. Security measures for key distribution and updates include: multi-layered encryption, security protocol protection, key encapsulation, rigorous authentication, and integrity checks to ensure keys safely reach their intended recipients, prevent man-in-the-middle attacks and data interception, and maintain the security foundation of the entire identity authentication system.

[0089] Example 2:

[0090] In this embodiment, establishing a hierarchical key management system based on the national cryptographic IBE framework includes: generating a system master key pair and selecting a secure elliptic curve base point through a root key generation center according to the SM2 elliptic curve algorithm and preset curve parameters to obtain a system-level master private key and master public key; calculating a dedicated domain master key for each identity domain through key derivation functions and SM3 hash operations based on the system-level master private key and the domain identifier of each identity domain, and establishing multiple domain key generation centers; and generating a user-specific identity private key and implementing the decentralized key distribution mechanism by mapping the user identity to a point on the elliptic curve through a double hash structure and performing scalar operations with the domain master key based on the user identity identifier and the domain master key.

[0091] Specifically, during the system initialization phase, the Root Key Generation Center (Root-PKG) is responsible for initializing the entire key system. This center first selects SM2 elliptic curve parameters conforming to Chinese national cryptographic standards, including key parameters such as the elliptic curve equation, base point G, and order n. Subsequently, the Root-PKG uses the SM2 algorithm to generate a system-level master key pair. The master private key msk is a randomly selected large integer within the range [1, n-1], and the master public key mpk is obtained through elliptic curve dot product operations, i.e., mpk = msk × G. This master key pair constitutes the root of trust for the entire key system, and its security directly affects the security of all lower-level keys.

[0092] The SM2 elliptic curve algorithm is an elliptic curve public-key cryptography standard and a core component of China's national cryptographic standards. Based on the mathematical principles of elliptic curves, the SM2 algorithm provides functions such as key pair generation, digital signatures, key exchange, and data encryption. The elliptic curve equation used in this algorithm is y² = x³ + ax + b (mod p), where a and b are curve coefficients, and p is a large prime number. In the national cryptographic standards, the curve parameters of the SM2 algorithm are predetermined, including the prime number p, coefficients a and b, the base point G, and its order n. These parameters undergo rigorous security analysis to ensure that the cryptographic strength meets national cryptographic management requirements.

[0093] An elliptic curve base point G is a specific point on the curve upon which all cryptographic operations are performed. The choice of base point G is crucial to the security of the algorithm; it must be ensured that the order n of G is a sufficiently large prime number, typically 256 bits or longer, to resist known cryptanalysis attacks. In the SM2 standard, the base point G and its parameters are already defined, and these preset parameters can be used directly. If custom parameters are required for specific scenarios, they must undergo rigorous cryptographic verification to ensure that their security is no less than that of the standard parameters.

[0094] The Root Key Generation Center (Root-PKG) is the root of trust in the entire key system, responsible for generating the system's master key pair. The system master key pair consists of the master private key (msk) and the master public key (mpk), and their generation process is as follows: First, the Root-PKG randomly selects an integer d as the master private key, with a value ranging from [1, n-1], where n is the order of the base point G; then, the master public key is calculated using the elliptic curve dot product operation P=d·G. The master private key must be generated and stored in a high-security hardware module (such as an encryption chip, cryptographic machine, etc.) to avoid the risk of leakage; the master public key, however, can be publicly released for use by various system components and users to verify identity.

[0095] A Key Derivation Function (KDF) is an algorithm that generates a purpose-specific key from the master key. It ensures that each derived key is independent of the others, meaning that even if one derived key is leaked, the security of the master key and other derived keys will not be affected. In this system, the KDF is used to derive the domain master keys for each identity field from the system's master private key. The SM3 hash algorithm is a cryptographic hash algorithm in the Chinese national cryptographic standard. It provides message digest functionality, maps input data of arbitrary length to a fixed-length (256-bit) output value, and has collision resistance and one-wayness, making it suitable for data processing during the key derivation process.

[0096] A Domain ID is a string that uniquely identifies a specific identity domain, typically containing information such as the organization name, department code, and business system identifier. Each identity domain is assigned a unique Domain ID to distinguish different Domain PKGs. The calculation process for the Domain Master Key is as follows: First, the Domain ID is hashed using the SM3 hash algorithm to obtain the Domain Feature Value H (DomainID); then, this Feature Value is mathematically multiplied (e.g., modular multiplication, scalar multiplication) with the system master private key to generate the domain's master private key; finally, the corresponding domain master public key is calculated based on the domain master private key. This derivation mechanism ensures that even if the master key of a certain domain is known, it is impossible to deduce the system master key or the master keys of other domains.

[0097] A Domain Key Generation Center (Domain-PKG) is an entity responsible for generating private keys for users within an identity domain. It exists within each identity domain and is operated and maintained by the domain administrator. Each Domain-PKG holds the domain master key assigned by the Root-PKG and has the authority to generate keys for users in its own domain, but cannot generate valid keys for users in other domains. The establishment process of a Domain-PKG includes: configuring dedicated hardware security devices to store the domain master key; deploying a key generation software system; establishing a user authentication mechanism; formulating a key management policy for the domain; and establishing a secure communication channel with the Root-PKG for key updates and state synchronization.

[0098] Double hashing is a commonly used method for generating user private keys in IBE systems, enhancing key security through two hash operations. In this system, the user private key generation process is as follows: First, the user's identity ID is mapped to a point P on an elliptic curve using the first hash function H1. ID = H1(ID); Then, use the second hash function H2 to perform a joint hash of the user's identity ID and the domain public key to generate a scalar value s = H2(ID, mpk domain Finally, calculate the user's private key dID = P. ID +msk domainThis double hash structure effectively prevents various known cryptographic attacks, ensuring the security and uniqueness of the user's private key.

[0099] The decentralized key distribution mechanism is implemented through a hierarchical PKG structure. Each domain PKG is responsible for key generation and distribution for users within its own domain, without the direct involvement of a central node. When a user needs to obtain a private key, they only need to submit authentication information and a key request to their domain's Domain-PKG. After verifying the user's identity, the Domain-PKG generates a unique private key for the user using the domain master key and the user's identity information, and distributes the private key to the user through a secure channel. This mechanism has several advantages: it reduces the load on the root PKG, avoiding the risk of single points of failure; it supports flexible intra-domain key management strategies, allowing each domain to formulate different update cycles and security policies according to its own needs; and it simplifies the cross-domain authentication process, as users only need to interact with their own domain PKG without needing to understand the complex structure of the entire system.

[0100] Example 3:

[0101] In this embodiment, the step of calculating a dedicated domain master key for each identity domain using a key derivation function and SM3 hashing includes: obtaining a domain feature value by hashing the domain identifier using an SM3-based hash function based on the system-level master private key and the domain identifier of each identity domain, thereby generating a unique identifier parameter for the domain; performing an elliptic curve operation on the domain feature value and the system master private key to generate a dedicated domain master private key for each identity domain; and calculating the corresponding domain master public key using an elliptic curve dot product operation based on the domain master private key and the elliptic curve base point, thereby establishing a domain-level key management system that supports the decentralized key distribution mechanism.

[0102] Specifically, the core of hierarchical key management lies in the establishment of a Domain Key Generation Center (Domain-PKG). Each identity domain represents a relatively independent security management area, which may correspond to different organizations, departments, or business systems. When a Domain-PKG needs to be established for a certain identity domain, the Root-PKG will generate a dedicated domain master key for that domain. The generation process of the domain master key uses a key derivation function, specifically calculated as follows: First, the domain identifier DomainID is hashed using the SM3-based hash function H3 to obtain the domain feature value. Then, this feature value is multiplied by the system master private key msk using an elliptic curve scalar multiplication operation to generate the master private key msk for that domain. domain Accordingly, the domain master public key mpk domain = msk domain × G. This derivation mechanism ensures that the key for each field is mathematically associated with the root key, while maintaining relative independence.

[0103] The SM3 hash function is a standard cryptographic hash algorithm. The SM3 algorithm can compress and map input messages of arbitrary length into a fixed-length (256-bit) output value, possessing cryptographic properties such as anti-preimage, anti-second preimage, and collision resistance. In the key derivation process, the SM3 hash function is used to process field identifiers, ensuring that even if field identifiers are similar, the generated field feature values ​​will be significantly different, enhancing the independence and security between fields.

[0104] A Domain ID is a string used to uniquely identify an identity domain. It typically contains multiple information fields, such as organization code, department number, and business type code, and may be formatted as "org.department.business" or other structured formats. Each identity domain must have a globally unique Domain ID to ensure that key materials between different domains are independent. During the initial setup phase, the system establishes a domain ID registration mechanism, with the security administrator responsible for allocating and managing domain IDs to ensure their uniqueness and compliance.

[0105] The domain signature is a fixed-length numerical value obtained by hashing the domain identifier, representing the cryptographic characteristics of the domain. The calculation process is as follows: taking the domain identifier DomainID as input, the SM3 hash function is used to calculate its hash value H = SM3(DomainID), yielding a 256-bit output. To accommodate elliptic curve operations, the system typically needs to convert this hash value into an integer within an appropriate range. Conversion methods may include modulo operations (taking the hash value modulo the curve order n) or truncation (taking a specific portion of the hash value). The resulting integer is the domain signature, used in subsequent domain master key generation.

[0106] Elliptic curve arithmetic refers to a series of computational operations based on the mathematical principles of elliptic curves, including point addition and point multiplication (scalar multiplication). In the process of generating the domain master private key, the main operations used by the system are scalar multiplication and modular arithmetic. Specifically, the system first converts the domain feature value h into an integer within an appropriate range, and then performs a modular multiplication with the system master private key msk: msk domain = (h The operation `msk) mod n`, where n is the order of the elliptic curve, ensures that the domain master private key is mathematically associated with the system master private key, while maintaining the uniqueness and unpredictability of each domain key.

[0107] The Domain Master Secret Key (DMK) is the core key material used by the Identity Domain Principal Module (PKG) to generate identity private keys for users within its domain. Each domain's DK is unique, held solely by the domain PKG, and must be stored and used in secure hardware. The security of the DK directly impacts the authentication security of all users in that domain; therefore, systems typically employ multi-layered protection measures, such as Hardware Cryptographic Module (HSM) storage, access control, operation auditing, and key fragmentation, to ensure the absolute security of the DK.

[0108] Elliptic curve product is a fundamental operation in public-key cryptography, referring to multiplying a point P on a curve by an integer k to obtain another point Q = k·P. In calculating the domain master's public key, the system uses the domain master's private key msk. domain As a scalar, perform a dot product with the base point G of the elliptic curve: mpk domain = msk domain ·G. Due to the computational difficulty of the elliptic curve discrete logarithm problem, even if the base point G and the domain master public key mpk are known... domain It is also impossible to derive the domain master private key msk within a feasible computational time. domain This ensures the security of the key system.

[0109] The Domain Master Public Key is a public parameter corresponding to the Domain Master Private Key, used to verify the validity of user private keys generated for that domain. The Domain Master Public Key can be securely and publicly distributed for use by system components and users. The computation result of the Domain Master Public Key is a point on an elliptic curve, typically represented as (x, y) coordinate pairs, or in a compressed format. In some IBE variants, the Domain Master Public Key may also include other auxiliary parameters, such as bilinear pairing parameters, to support more complex cryptographic operations.

[0110] The domain-level key management system serves as the middle layer of the entire hierarchical key management architecture, connecting the root key management center and user key management. It is a crucial link in achieving decentralized key distribution. This system comprises multiple parallel-running domain key generators (PKGs), each with independent key generation capabilities and responsible for managing the key lifecycle of users within its identity domain. The main functions of the domain-level key management system include: domain-wide user authentication; user private key generation and distribution; key updates and revocations; enforcement of domain-wide access control policies; and synchronous communication with the root PKG. Through this hierarchical management model, the system achieves a reasonable allocation of key management tasks, maintaining consistency in global policies while supporting the flexibility of local management, providing an efficient and reliable key infrastructure for large-scale identity authentication systems.

[0111] Example 4:

[0112] In this embodiment, the construction of the domain-aware differential privacy protection module involves obtaining security level information of the user identity domain, defining a privacy budget allocation strategy based on the security level information, and adding calibrated Laplace noise to the user identity feature vector to generate privacy-protected identity authentication data. This includes: based on the decentralized key distribution mechanism and the data sensitivity of each identity domain, analyzing the privacy requirements of information within the domain using a security level assessment algorithm to determine a tiered privacy budget allocation strategy; based on the privacy budget allocation strategy and the user identity feature vector, calculating function sensitivity and generating calibrated Laplace noise to obtain identity feature data with added privacy protection noise; and based on the identity feature data with added privacy protection noise, implementing collaborative training of identity recognition models across domains using a heterogeneous joint learning framework, and employing gradient compression and secure aggregation protocols to protect the model parameter exchange process to obtain the privacy-protected identity feature data.

[0113] Specifically, constructing a domain-aware differential privacy protection module is a core step in achieving privacy-preserving identity authentication. Domain-aware differential privacy is an enhancement and extension of traditional differential privacy, providing a refined privacy protection mechanism by considering the characteristics and security requirements of different identity domains. An identity domain refers to a group of users with similar attributes or belonging to the same organizational unit; different identity domains may face different levels of privacy risks and protection needs. The domain-aware approach allows the system to adjust the strength of privacy protection according to the specific circumstances of each domain, maximizing data availability while protecting user privacy.

[0114] Differential privacy is a rigorous mathematical framework for privacy protection, providing provable privacy guarantees for data analysis. Its core idea is to add carefully calibrated random noise during data processing, making it impossible to determine from the analysis results whether any target user is included in the dataset. The formal definition of differential privacy is based on the concept of neighboring datasets: if an algorithm, when run on any pair of datasets differing by only one record, exhibits output distributions whose differences are strictly bounded, then the algorithm satisfies differential privacy. This bound is controlled by the privacy budget (ε) parameter; a smaller ε value provides stronger privacy protection, but may reduce the usefulness of the data.

[0115] Obtaining security level information for user identity domains is the first step in implementing differential privacy protection. The system first identifies and defines different identity domains, such as public domains, internal domains, sensitive domains, and core domains. A security level assessment is performed on each domain, considering the following factors: data sensitivity (assessing the potential harm caused by data breaches); legal compliance requirements (different types of data may be protected by different regulations); threat model (assessing the capabilities and motivations of potential attackers); user expectations (considering users' reasonable expectations for privacy protection); and data usage purpose (balancing privacy protection with functional requirements). The assessment results form a domain security level matrix, providing a basis for subsequent privacy budget allocation. The specific implementation method for security level assessment of each identity domain is as follows: Basic information for each domain is collected through a pre-set assessment module, including but not limited to a list of data types stored or processed within the domain (such as whether it contains biometric information, financial data, personal identification information, etc.), a list of applicable relevant regulations, historical security event records, user privacy preference settings data, and business data flow diagrams. The assessment module assigns weights and scores to the above factors based on a pre-set rule engine and quantitative model. For example, five levels (1-5) are set for data sensitivity, with higher levels indicating greater harm from data breaches. The score is determined based on the most sensitive data type within the domain. Scores for legal compliance requirements are set based on the severity of penalties for violations. Threat models are quantitatively assessed based on publicly available threat intelligence and the historical attack frequency and severity of the domain. Finally, a weighted summation algorithm is used to calculate a quantitative security level score for each domain, and the score is mapped to a predefined security level (e.g., public domain, internal domain, sensitive domain, core domain). The assessment results, including the scores, levels, and key influencing factors for each domain, are stored and updated in a domain security level matrix. This matrix is ​​a data structure with domain IDs as rows and each assessment factor and the total score as columns, which can be directly read and used by the subsequent privacy budget allocation strategy module.

[0116] Data sensitivity assesses the potential harm caused by a data breach. Specifically, data sensitivity is a quantitative indicator of the potential harm caused by a data breach, defined and obtained based on a pre-established data sensitivity grading system. A data sensitivity grading rule base is maintained, detailing the sensitivity levels corresponding to different data types. For example, biometric data (such as fingerprints and iris information) might be defined as the highest sensitivity level (e.g., level 5); personal identification information (such as ID card numbers and mobile phone numbers) might be defined as the second highest level (e.g., level 4); social relationship data and behavioral pattern data might be defined as medium sensitivity (e.g., level 3); and general business operation logs might be defined as low sensitivity (e.g., levels 1-2). When assessing the data sensitivity of a specific identity domain, all data types processed within that domain are scanned and identified. The sensitivity level of each data type is mapped according to the rule base, and the highest level is taken as the overall data sensitivity assessment result for that domain. The assessment of the severity of data breaches is conducted through a pre-defined risk analysis model. This model considers factors including the direct impact of sensitive data breaches (such as identity theft and financial loss), indirect impacts (such as reputational damage and business interruption), and the scale of affected users. This model correlates the sensitivity level with the specific severity of the breach, providing a basis for security level scoring.

[0117] Decentralized key distribution mechanisms are the infrastructure supporting domain-aware privacy protection. This mechanism employs a distributed architecture, avoiding single points of failure and the risks of centralized control. Each identity domain maintains its own key management node, responsible for generating and distributing key materials used for encryption and noise generation. Nodes exchange necessary information through secure communication protocols, but each maintains independent decision-making power. The decentralized design allows different domains to implement customized privacy protection strategies based on their own characteristics and needs, while a consensus protocol ensures the overall system's consistency. This architecture enhances system security and scalability, making it particularly suitable for cross-organizational identity authentication scenarios.

[0118] Security level assessment algorithms are structured methods used to quantitatively evaluate the privacy protection needs of different identity domains. The algorithm considers multi-dimensional factors, including: data type sensitivity. Different types of identity data (such as biometrics, social relationships, and behavioral patterns) have different sensitivities. Sensitivity specifically refers to the degree to which each dimension of the user's identity feature vector is sensitive to revealing the user's true identity information. Its definition is based on the strength of the association between that dimension and the user's unique identifier and the level of privacy risk. Methods for obtaining and defining the sensitivity values ​​of each dimension include: 1) Quantitative analysis based on privacy risk models, where the system calculates the probability that an attacker could successfully infer or forge a user's identity if they knew the precise value of that dimension; the higher the probability, the higher the sensitivity; 2) Referencing the classification requirements for different types of identity data in domestic and international privacy protection standards and related regulations; 3) Manual evaluation and labeling by security experts. For example, biometric dimensions are assigned the highest sensitivity value due to their high uniqueness and immutability; while commonly used device fingerprint information has a relatively low sensitivity value. During initialization or model updates, a specific sensitivity value is assigned to each dimension of the user identity feature vector based on the above methods. This value is stored in a feature sensitivity configuration table and used for subsequent calculation of differential privacy noise parameters. The assessment considers factors such as data usage scope (data with a wider usage scope requires stronger protection), regulatory compliance requirements (meeting specific requirements of privacy regulations such as GDPR and CCPA), historical security incidents (considering lessons learned from past data breaches), and threat intelligence (using the latest threat intelligence to assess the risk level). The assessment process uses a weighted scoring model, standardizing and weighting each factor to output a security level score from 0 to 10, with higher scores indicating stronger privacy protection requirements.

[0119] A tiered privacy budget allocation strategy is a method for rationally allocating a limited overall privacy budget to different identity domains and data attributes. The privacy budget is a crucial parameter for measuring the strength of privacy protection; a smaller budget generally indicates stronger privacy protection, but may reduce data usability. The allocation strategy is based on the following principles: risk proportion principle, allocating more budget to high-risk domains (i.e., smaller ε values); data value balance, considering the analytical value of data against privacy risks; use case adaptation, adjusting the budget according to the data's usage scenario; and dynamic adjustment capability, allowing for adjustments to budget allocation based on changes in the threat landscape. In practice, a base budget is first set for each domain, then adjusted according to security levels and requirements, ultimately ensuring that the total budget usage does not exceed the system's set global limit.

[0120] A user identity feature vector is a multi-dimensional numerical representation of a user's identity attributes. These features may include biometrics (such as facial features and fingerprint patterns), behavioral features (such as typing patterns and operating habits), and knowledge features (such as password usage patterns). Feature vectors are typically normalized to ensure that the values ​​of each dimension are within a similar range, facilitating subsequent processing. Regarding privacy protection, different feature dimensions may have different sensitivities, requiring differentiated protection. Maintaining a feature sensitivity mapping table guides the noise addition process. The construction of feature vectors needs to balance recognition accuracy and privacy protection requirements, selecting a feature set that is both discriminative and does not excessively expose personal privacy.

[0121] Function sensitivity is a key concept in differential privacy mechanism design, used to quantify the maximum responsiveness of a function's output to changes in a single input. Intuitively, sensitivity measures the maximum possible change in the function's output when a record is added or removed. For authentication scenarios, it's necessary to calculate the sensitivity of various feature extraction and matching functions to ensure an appropriate amount of noise is added. Sensitivity calculation methods include: theoretical analysis, determining the upper bound of the function's sensitivity through mathematical derivation; empirical estimation, estimating sensitivity by testing the function's behavior on sample data; and adaptive calculation, dynamically adjusting the sensitivity estimate based on the actual data distribution at runtime. Accurate sensitivity calculation is fundamental to ensuring the effectiveness of differential privacy protection; overestimation leads to excessive noise affecting usability, while underestimation may weaken privacy protection.

[0122] Laplace noise is a commonly used type of random noise for achieving differential privacy, with its probability density function exhibiting a centrally symmetric, exponentially decaying form. The choice of the Laplace distribution is based on its mathematical properties, making it particularly suitable for meeting the requirements of differential privacy. Calibrated Laplace noise refers to the amount of noise precisely adjusted according to the function sensitivity and a preset privacy budget, ensuring that the data after adding noise meets the definition of differential privacy. The noise generation process includes: determining the target privacy budget ε; calculating the function sensitivity Δf; generating random noise according to the Laplace distribution, with a scaling parameter of Δf / ε; and adding the generated noise to each dimension of the original feature vector. To accommodate differences in sensitivity across different dimensions, the system may apply different proportions of noise to different feature dimensions.

[0123] The noise addition process requires meticulous control to ensure that privacy is protected without unduly impacting the accuracy of identity authentication. The system implements the following techniques to improve the effectiveness of noise addition: dimensionality sensitivity analysis to identify feature dimensions that significantly affect authentication accuracy and allocate them with smaller noise levels; noise calibration optimization to find the optimal noise parameter configuration through iterative testing; boundary processing to ensure that the data after noise addition remains within a valid range; correlation preservation to consider the correlation between features when adding noise, avoiding disruption of important feature relationships; and post-processing techniques to apply specialized algorithms to mitigate the impact of noise on subsequent authentication processes. These techniques collectively ensure that the identity feature data after adding privacy-preserving noise still possesses sufficient identifiable value.

[0124] Heterogeneous joint learning frameworks are distributed machine learning architectures that allow multiple participants to collaboratively train models without sharing the original data. In identity authentication scenarios, this framework enables different identity domains to jointly build more powerful recognition models while protecting the privacy of their respective data. Heterogeneous refers to the fact that participants may have different data distributions, system architectures, and computing capabilities, and the framework needs to handle this diversity. The basic process of joint learning includes: local model training, where each participant trains the model using local data; model parameter exchange, where participants only share model parameters (such as gradients) rather than the original data; parameter aggregation, where parameters from all parties are aggregated to update the global model; and model distribution, where the updated global model is distributed to all participants. The entire process achieves a privacy-preserving paradigm of "data remains unchanged, model moves."

[0125] Gradient compression is a crucial technique for optimizing communication efficiency in joint learning. It reduces bandwidth requirements and potential privacy risks by decreasing the amount of transmitted gradient information. Key compression techniques include: quantization, mapping gradient values ​​to a finite set (e.g., converting a 32-bit floating-point number to an 8-bit integer); sparsity, transmitting only important gradient values ​​and ignoring small gradients close to zero; random masking, randomly selecting a subset of gradients for transmission; and low-rank decomposition, decomposing the gradient matrix into a low-rank form to reduce the number of parameters. The compression process needs to balance communication efficiency and model performance; excessive compression can lead to training instability or slower convergence. The system minimizes the impact of compression on model training by dynamically adjusting the compression ratio and employing an error feedback mechanism.

[0126] Secure aggregation protocols are cryptographic mechanisms that protect the parameter exchange process in joint learning, ensuring that no participant or external observer can infer the original data of other participants from the exchanged parameters. Based on the principle of secure multi-party computation, the protocol's main implementation methods include: homomorphic encryption, allowing parameter aggregation in an encrypted state; secret sharing, distributing parameters into multiple shares, requiring a sufficient number of shares to reconstruct the original values; differential privacy aggregation, adding noise during the aggregation process to protect individual contributions; and secure multi-party computation, using a special protocol to compute function results while protecting input privacy. The protocol implementation considers a balance between communication efficiency, computational overhead, and security strength, and is optimized for the characteristics of authentication scenarios, ensuring that privacy is protected without unduly impacting model training performance.

[0127] By implementing domain-aware differential privacy protection, a successful authentication mechanism capable of protecting user identity privacy has been constructed. The core advantages of this mechanism are: providing differentiated protection based on the characteristics and needs of different identity domains; mathematically provable privacy guarantees; maintaining data availability through precisely calibrated noise; and supporting secure collaboration between different domains to build more robust recognition models. This approach provides strong privacy protection for identity authentication in sensitive scenarios, enabling organizations to comply with increasingly stringent data protection regulations while ensuring security. The entire process achieves an effective balance between privacy protection and identity authentication functions, providing crucial support for building a trusted digital identity infrastructure.

[0128] Example 5:

[0129] In this embodiment, the step of calculating function sensitivity and generating calibrated Laplace noise includes: determining the function sensitivity values ​​of various identity features based on the privacy budget allocation strategy and each dimension of the user identity feature vector through theoretical analysis and experimental verification, and obtaining noise scale calculation parameters. The theoretical analysis involves mathematically deriving the upper bound of the global sensitivity Δf for the identity feature extraction or processing function f. For example, for a function calculating the norm of a feature vector, the sensitivity is the maximum possible norm difference within the range of feature value changes; for functions involving counting, the sensitivity is typically 1. The experimental verification is based on the theoretical analysis and uses historical identity feature datasets for simulation verification. The specific method includes: randomly selecting a pair of neighboring datasets D and D' that differ by only one record from the dataset; calculating the outputs f(D) and f(D') of the function f on D and D'; calculating a measure of the output difference (such as the L1 or L2 norm); repeating this process a large number of times, taking the maximum value of all difference measures or a high percentile (such as the 95th percentile) as the experimentally observed sensitivity estimate; finally determining the function sensitivity value, taking the larger of the theoretical analysis upper bound and the experimental verification estimate to ensure the strength of privacy protection; generating calibrated Laplace noise according to the ratio of the noise scale calculation parameter to the privacy budget parameter in the privacy budget allocation strategy; adding the calibrated Laplace noise to each dimension of the user identity feature vector through vector addition to obtain the identity feature data with added privacy protection noise.

[0130] Specifically, calculating function sensitivity is a fundamental step in achieving differential privacy. Function sensitivity is defined as the maximum possible response of a function's output to changes in its input; it quantifies the maximum change in the function's output when a single data record changes. In authentication systems, function sensitivity measures the maximum possible change in the authentication function's result when a user's identity characteristics undergo minor changes. Accurately calculating function sensitivity is crucial for determining the appropriate amount of noise; an excessively low sensitivity estimate leads to insufficient privacy protection, while an excessively high sensitivity introduces unnecessary noise and reduces system accuracy.

[0131] Theoretical analysis is the primary method for determining the sensitivity of a function. For different dimensions of the identity feature vector, the system uses mathematical derivation to analyze the influence range of each feature in the authentication process. For linear functions (such as Euclidean distance calculation), sensitivity equals the maximum possible change in the feature vector; for nonlinear functions (such as biometric matching algorithms), it is necessary to analyze the upper bound of the function's gradient. Theoretical analysis considers the following factors: the range of feature values ​​to determine the maximum possible change in the features; the mathematical properties of the function to analyze its response characteristics to input changes; and extreme case analysis to identify boundary conditions that may lead to the maximum output change. Theoretical analysis provides an upper bound estimate of sensitivity, offering theoretical assurance against noise addition.

[0132] Experimental validation is an important means of supplementing theoretical analysis. Using a large amount of real and simulated identity data, the actual changes in the function output when a single feature changes are tested. Experimental validation includes the following steps: constructing a test dataset containing diverse identity feature samples; generating neighboring datasets by making minor modifications to individual records in the original dataset; executing the objective function, running the authentication function on both the original and modified datasets; measuring the output differences, calculating and recording the amount of change in the output results; and statistical analysis to determine the maximum observed change and distribution characteristics. Through extensive experimentation, more accurate sensitivity estimates can be obtained than theoretical analysis, especially for complex nonlinear authentication functions.

[0133] The noise scale calculation parameter is a specific value determined based on the function sensitivity, directly determining the amount of noise to be added. Different sensitivity values ​​may be determined for different dimensions of the user identity feature vector to reflect the varying importance of each dimension in the authentication process. A high sensitivity value indicates that the feature has a significant impact on the authentication result, requiring more noise protection; a low sensitivity value indicates a smaller impact, allowing for the addition of less noise. A sensitivity mapping table is maintained to store the corresponding sensitivity value for each feature dimension. These values ​​combine the results of theoretical analysis and experimental verification, ensuring sufficient privacy protection while preserving as much data usefulness as possible.

[0134] The privacy budget parameter (ε) is a key parameter in differential privacy mechanisms that controls the strength of privacy protection; it is also known as the privacy loss parameter. A smaller ε value provides stronger privacy protection, but may reduce data usability; a larger ε value provides better data usability, but weakens privacy protection. Based on the aforementioned privacy budget allocation strategy, different ε values ​​are assigned to different identity domains and feature dimensions to reflect their varying privacy protection needs. For highly sensitive biometric data, a smaller ε value (e.g., 0.1) may be assigned; for relatively insensitive usage habit data, a larger ε value (e.g., 1.0) may be assigned. This differentiated allocation satisfies the protection needs of different data types while optimizing overall data usability.

[0135] Laplace noise is a standard technique for achieving differential privacy, characterized by its simplicity, ease of implementation, and sound mathematical properties. The Laplace distribution is a symmetric exponential probability distribution whose probability density function reaches its maximum at the center and decays exponentially with increasing distance from the center. In differential privacy applications, the magnitude of the Laplace noise is controlled by a scaling parameter b, which is equal to the ratio of the function sensitivity to the privacy budget (Δf / ε). This design ensures that the data with added noise meets the requirements of ε-differential privacy, providing a provable level of privacy protection.

[0136] The calibrated Laplace noise generation process includes the following steps: First, for each feature dimension i, calculate its noise scaling parameter bi = Δfi / εi, where Δfi is the function sensitivity of that dimension, and εi is the privacy budget allocated to that dimension. Then, using a standard random number generator, generate random noise conforming to a Laplace distribution for each dimension, with the scaling parameter being the corresponding bi. Finally, perform a range check on the generated noise to ensure that the data after adding noise remains within a valid range, and truncate it if necessary. This dimension-calibrated noise generation method ensures differentiated protection for features with different sensitivities, optimizing the balance between privacy protection and data usefulness.

[0137] Vector addition is the process of adding generated noise to the original identity feature vector. For an n-dimensional feature vector X=(x1,x2,...,x...),... n ) and the corresponding generated noise vector N=(n1,n2,...,n n The result of adding noise is X'=(x1+n1,x2+n2,...,x n +n n This simple vector addition ensures that each dimension of the original data is protected from noise. To maintain data validity, post-processing is performed after addition: range constraints to ensure that the values ​​after noise addition remain within the valid range; anomaly detection to identify and adjust extreme noise that may lead to anomalous results; and consistency checks to maintain necessary relationship constraints between features. The simplicity of vector addition makes it easy to implement and computationally efficient, suitable for the real-time processing needs of large-scale identity authentication systems.

[0138] Through the above steps, identity feature data with added privacy-preserving noise was successfully generated. This data provides strong privacy protection while retaining sufficient recognizability. The differential privacy mechanism ensures that even if an attacker possesses all data except for the target user, they cannot accurately infer the target user's true characteristics, thus protecting the user's identity privacy. This method provides mathematically provable privacy guarantees for identity authentication in sensitive environments, balancing the multiple requirements of security, privacy protection, and system availability.

[0139] Example 6:

[0140] In this embodiment, the design of the distributed batch matrix multiplication protocol, which obtains key update requests from multiple users and decomposes them into multiple computing nodes for parallel processing using secret sharing technology to generate batch-updated user keys, includes: based on the privacy-protected identity authentication data and the key update requests from multiple users, converting serial key updates into batch matrix operations by constructing a user identity matrix and a key generation matrix to obtain a matrix-based key generation task; according to the matrix-based key generation task, decomposing the user identity matrix and the key generation matrix into multiple matrix fragments using the Shamir secret sharing scheme and distributing them to different computing nodes to achieve distributed storage of sensitive information; based on the matrix fragments of each computing node, calculating the multiplication results of each fragment by performing local matrix multiplication operations in parallel and reconstructing the final result using Lagrange interpolation to generate the batch-updated user keys.

[0141] Specifically, the distributed batch matrix multiplication protocol is a secure computational method for efficiently handling large-scale key update requests. This protocol leverages the parallel nature of matrix operations and the security features of secret sharing technology to achieve batch key generation and updates while protecting user privacy. Traditional key updates typically employ a serial processing approach, handling user requests one by one, which is inefficient when faced with a large number of concurrent requests. Batch processing significantly improves computational efficiency by transforming multiple independent key update operations into unified matrix operations, while simultaneously enhancing system scalability and security through a distributed computing architecture.

[0142] The user identity matrix is ​​a two-dimensional data structure representing the identity features of multiple users. Each row of the matrix corresponds to a user's identity feature vector, and each column corresponds to a different dimension of the feature. For n users, each user has m-dimensional features, and the user identity matrix is ​​an n×m matrix. When constructing this matrix, the system first collects the identity data of all users whose keys need to be updated, including identity features protected by differential privacy; then, these feature vectors are integrated into a matrix according to a predetermined format; finally, necessary normalization and standardization processing is performed to ensure that the values ​​of different dimensions are within a comparable range. The user identity matrix contains highly sensitive information, and its secure storage and processing are core considerations in the system design.

[0143] The key generation matrix is ​​a transformation matrix used by the system to calculate user keys. This matrix encodes the mapping relationship from user identity characteristics to key material, and its construction is based on cryptographic methods such as the aforementioned key derivation function and SM3 hash operation. For a user with m-dimensional characteristics and an l-bit key to be generated, the key generation matrix is ​​an m×1 matrix. Each element of this matrix may originate from the parameters of the key derivation function, a random seed, or a derived value of the system master key. The design of the key generation matrix must ensure that: different users generate different keys; small changes in user characteristics do not lead to drastic changes in the key; the key has sufficient entropy and unpredictability; and even if some matrix elements are leaked, it will not significantly reduce key security.

[0144] Matrix-based key generation refers to representing the key generation process as matrix multiplication. Traditional key updates require a series of independent cryptographic operations for each user, while matrix-based methods allow simultaneous key computation for multiple users through a single matrix multiplication. Specifically, the user identity matrix (n×m) is multiplied by the key generation matrix (m×l) to obtain a key result matrix (n×l), where each row corresponds to a user's new key. This transformation leverages the high optimization of matrix operations in modern computing systems, particularly on GPUs and dedicated hardware accelerators, significantly improving computational efficiency. Furthermore, matrix-based processing facilitates the application of cryptographic protection techniques, such as homomorphic encryption and secure multi-party computation, enhancing the security of the key generation process.

[0145] The Shamir secret-sharing scheme is a classic threshold cryptography technique that allows secret information to be broken down into multiple shares. Only when a sufficient number of shares are collected can the original secret be reconstructed. This scheme is based on the principle of polynomial interpolation. Specifically: a threshold *t* is chosen, representing the minimum number of shares required to reconstruct the secret; a polynomial of degree *t-1* is constructed, with the constant term being the secret to be shared and other coefficients randomly selected; *n* distinct points are selected on this polynomial to generate *n* shares; only when at least *t* shares are collected can the polynomial be reconstructed through interpolation, thus obtaining the secret value. The security of this scheme is based on the fact that when fewer than *t* shares are known, the secret can be any value, possessing information-theoretic security. In this system, the Shamir scheme is used to share matrix elements, ensuring that no single computing node can obtain the complete sensitive information.

[0146] Matrix sharding is the process of decomposing the entire matrix into multiple parts, so that different computing nodes only hold partial information. The system applies Shamir's secret sharing to each element of the user identity matrix and key generation matrix, generating multiple matrix shards. Specific steps include: determining sharing parameters, including the number of computing nodes *n* and the reconstruction threshold *t*; applying the Shamir scheme to each matrix element to generate *n* shares; constructing *n* matrix shards based on the share distribution, each shard containing one share of all elements in the original matrix; and securely distributing the shards to the corresponding computing nodes. This sharding method ensures that even if some nodes are compromised, attackers cannot obtain complete sensitive information, while also resisting a certain proportion of node failures, thus improving system reliability.

[0147] Distributed storage refers to distributing sensitive information across multiple physically isolated computing nodes to prevent single-point leaks from causing overall security risks. In this system, each computing node stores only matrix fragments, not the complete matrix, thus implementing the security principle of "distributed knowledge." The implementation of distributed storage considers the following aspects: node selection, choosing nodes with different security domains and management permissions to reduce the risk of collusion; secure communication, using encrypted channels to transmit fragmented data; access control, strictly limiting node access permissions to fragments; integrity protection, using digital signatures to ensure fragments are not tampered with; and high availability design, deploying redundant nodes to cope with possible node failures. This distributed architecture significantly improves the system's ability to resist internal threats and external attacks.

[0148] Parallel matrix multiplication is the core operation performed locally on each computing node. Each node holds a fragment of the user identity matrix and the key generation matrix, and performs local multiplication calculations on these fragments. Due to the nature of matrix multiplication, nodes can compute partial results without knowing the complete matrix. The parallel computing steps include: node initialization, loading fragment data and preparing the computing environment; local computation, performing multiplication operations on the fragmented matrices; result processing, applying necessary post-processing to the computation results; and result transmission, securely sending the partial results to the result aggregator. The system fully leverages the parallel processing capabilities of modern hardware, such as multi-core CPUs, GPUs, and dedicated computing accelerators, to further improve computational efficiency.

[0149] Lagrange interpolation is a mathematical tool for reconstructing secret-sharing results. Once enough computational node results are obtained, the system uses a Lagrange interpolation polynomial to reconstruct the final key result matrix. The specific steps include: collecting the computational results from each node, with each result representing a share of the final matrix; applying the Lagrange interpolation formula to each matrix position to reconstruct the original value based on multiple shares; verifying the correctness of the reconstructed results to ensure no node submitted incorrect or malicious data; and assembling the complete result matrix, containing the new keys for all users. Lagrange interpolation provides a mathematically rigorous reconstruction method, ensuring that the original result can be accurately reconstructed as long as a sufficient number of correct shares are collected.

[0150] The batch-updated user keys are the final output of the protocol execution. In the reconstructed matrix, each row corresponds to a new key for a user. The system performs the following post-processing steps: key formatting, converting matrix elements to a standard key format; key verification, ensuring the generated keys meet strength and uniqueness requirements; secure distribution, transmitting the new keys to the corresponding users through a secure channel; key update recording, updating key status information in the system and recording audit logs; and old key processing, arranging appropriate old key transition and revocation strategies. Batch processing significantly improves the system's ability to handle a large number of concurrent key update requests while maintaining the security and privacy of the key generation process.

[0151] The system achieves efficient and secure large-scale user key updates through a distributed batch matrix multiplication protocol. This method distributes the computational load across multiple nodes, improving both processing speed and security, as no single node can access all sensitive information. Secret-sharing technology ensures that even if some nodes are compromised, the overall system security remains intact. This approach, combining high-performance computing and cryptographic protection, provides robust key management capabilities for large-scale identity authentication systems.

[0152] Example 7:

[0153] In this embodiment, the implementation of the zero-knowledge proof verification mechanism involves constructing a zero-knowledge proof protocol using a commitment scheme based on the SM3 hash algorithm. This process completes identity verification and generates an authentication result without exposing the user's private key. The steps include: generating a corresponding public key based on the batch-updated user key and preset elliptic curve parameters using discrete logarithm operations and storing it on an authentication server, establishing a public-private key mapping; calculating a commitment value and receiving a random challenge from the verifier based on a randomly selected temporary value, the elliptic curve parameters, and the user's private key in the public-private key mapping, generating a response value containing the user's private key information; and verifying the correctness of the zero-knowledge proof equation using the SM3 hash algorithm based on the commitment value, the random challenge, and the response value, thus completing identity verification and generating the authentication result without exposing the user's private key.

[0154] Specifically, zero-knowledge proof verification is an advanced cryptographic technique that allows one party (the prover) to prove the truth of a statement to another party (the verifier) ​​without revealing any additional information beyond the statement's truthfulness. In identity authentication scenarios, this means a user can prove they possess the correct private key without actually revealing it. Zero-knowledge proofs have three fundamental properties: completeness (if the statement is true, an honest verifier will be convinced); soundness (if the statement is false, any deception strategy has only a negligible probability of success); and zero-knowledge (the verifier gains no additional information beyond the statement's truthfulness). This mechanism effectively addresses the privacy risks of traditional authentication methods and is particularly suitable for identity authentication scenarios with high security requirements.

[0155] SM3 compresses input messages of arbitrary length into a fixed-length (256-bit) output value, exhibiting cryptographic properties such as one-wayness, collision resistance, and avalanche effect. The main processing steps of the algorithm include: message padding, filling the message to a multiple of 512 bits; message expansion, generating extended message blocks for compression; compression function, iteratively processing the extended message blocks and updating the linker variables; and output transformation, outputting the final linker variables as hash values. In zero-knowledge proof systems, SM3 is used to generate commitment values, handle challenges, and verify responses, providing cryptographic security guarantees. SM3's domestically developed nature ensures the system complies with domestic cryptographic regulations, facilitating regulatory approval and deployment in official and critical infrastructure environments.

[0156] Elliptic curve parameters are the set of fundamental elements defining an elliptic curve cryptosystem. Elliptic curve cryptography (ECC) is based on the algebraic structure of point groups on an elliptic curve, requiring shorter key lengths compared to traditional cryptosystems like RSA while maintaining the same level of security. Predefined elliptic curve parameters typically include: curve equation parameters a and b, defining the shape of the curve; finite field parameter p, determining the prime field of operation; a base point G, a point on the elliptic curve that serves as the starting point for public key generation; the order of the base point n, the smallest positive integer n such that nG equals infinity; and a cofactor h, the ratio of the total number of points on the curve to the order of the base point. The system may use an international standard curve (such as NIST P-256) or a national standard curve (such as the SM2 curve). The choice of curve must balance security and performance, ensuring resistance to known elliptic curve cryptanalysis attacks.

[0157] Discrete logarithm operation is a core operation in elliptic curve cryptography (ECC), used to generate a corresponding public key from a private key. On an elliptic curve, the private key is a large integer *d*, and the public key is a point on the curve, P = d × G, where G is a predetermined base point, and * ×* represents scalar multiplication of the point (adding G d times). The security of this operation is based on the computational difficulty of the Elliptic Curve Discrete Logarithm Problem (ECDLP): given the public key P and the base point G, calculating the private key *d* is computationally infeasible. When performing this operation, the following aspects must be considered: optimization for large number operations to efficiently handle integers of hundreds of digits; selection of dot multiplication algorithms, such as double addition, windowing, or the Montgomery step algorithm; side-channel protection to prevent leakage of private key information through side channels such as time and power consumption; and parameter verification to ensure that the generated public key indeed lies on the specified elliptic curve. The irreversibility of discrete logarithm operation is the foundation of the security of authentication systems.

[0158] The public-private key mapping establishes a mapping between each user's private key and their publicly disclosed public key. The private key, generated through the aforementioned batch update process, is sensitive information known only to the user; the public key, derived from the private key through discrete logarithm operations, can be securely published. The steps for establishing the public-private key mapping include: for each user, reading their updated private key; performing a dot product operation using preset elliptic curve parameters and the private key to calculate the coordinates of the public key points; converting the public key to a standard format (such as compressed or uncompressed); establishing a mapping between the user identifier and the public key; and storing the mapping in the secure database of the authentication server. Appropriate access control and encryption protection measures are implemented during storage to prevent unauthorized access. This mapping enables the system to verify user identity while maintaining the confidentiality of the private key.

[0159] Commitment schemes are a fundamental component of zero-knowledge proofs, allowing the prover to "commit" to a value without immediately revealing it. Commitment schemes based on SM3 hashing offer binding (the value cannot be changed after commitment) and concealment (the committed value is not disclosed). In this system, the user (prover) generates a commitment value through the following steps: selecting a random temporary value r as a confusion factor; calculating a temporary point T = r × G, where G is the base point of an elliptic curve; using the coordinates of the temporary point T as the commitment value C, or further calculating C = SM3(T) using SM3 hashing; sending the commitment value C to the verifier, but keeping r and the private key d hidden. The commitment value C indicates to the verifier that the user has "locked" a specific value, and subsequent steps will prove that the value matches the user's claimed identity, without revealing the private key itself. The security of the commitment scheme directly relies on the randomness of the temporary value r and the cryptographic properties of SM3 hashing.

[0160] The random challenge is a random value generated and sent to the prover by the verifier after receiving the commitment value. This step introduces unpredictability, ensuring that the prover cannot prepare a response in advance. The verifier's process for generating the random challenge includes: using a cryptographically secure random number generator; ensuring the challenge value e has an appropriate bit length and range; sending the challenge value e to the prover; and recording the challenge value for subsequent verification. The random challenge can be generated directly using random numbers or derived by applying SM3 hashes to relevant information (such as the commitment value, user identifier, timestamp, etc.). The randomness of the challenge is crucial to the protocol's security, preventing replay attacks and pre-computed attacks. In practical implementations, the size of the challenge value may be limited or its format may be pre-defined to balance security and efficiency.

[0161] The response value is the proof calculated by the prover based on the private key, a temporary random number, and the received challenge. The formula for calculating the response value is typically s = r + e × d (mod n), where r is the previously chosen temporary random number, e is the received challenge, d is the user's private key, and n is the order of the elliptic curve base point. Key characteristics of the response value s are: it contains private key information, but due to the inclusion of the random number r, the private key d cannot be extracted from s; and it has a verifiable mathematical relationship with the previously sent commitment value and public key. When generating the response value, attention should be paid to: correct implementation of modulo arithmetic to ensure the result is within an appropriate range; avoiding the use of insecure random number generators to prevent private key leakage; ensuring the response value does not reveal statistical information about the private key; and implementing necessary side-channel protection measures to prevent physical attacks. The response value is the core output of the proof process, proving both the possession of the private key and protecting its confidentiality.

[0162] Zero-knowledge proof equations are mathematical relations used by the verifier to check the validity of a proof. Upon receiving the response value s, the verifier performs the following verification: calculates the point R = s × Ge × P, where G is the base point, P is the user's public key, and e is the previous challenge; checks whether R equals the ephemeral point T in the initial commitment, or whether SM3(R) equals the commitment value C. This verification utilizes an algebraic relation on an elliptic curve: if s = r + e × d and P = d × G, then s × Ge × P = (r + e × d) × Ge × (d × G) = r × G = T. The validity of the verification equation proves that the user does indeed know the private key d, without the user directly revealing the private key. The entire verification process uses only public information (public key, commitment value, challenge, and response), requiring no secret input, embodying the core characteristic of zero-knowledge proofs.

[0163] The authentication process is a complete interactive flow of zero-knowledge proof. The interaction between the user (provider) and the system (verifier) ​​includes three basic steps: In the commitment phase, the user generates a temporary random number r, calculates the commitment T = r × G, and sends the commitment value C to the verifier; in the challenge phase, the verifier generates a random challenge e and sends it to the user; in the response phase, the user calculates the response s = r + e × d and sends it to the verifier, who checks if the equation s × Ge × P = T holds true. If the verification passes, the system confirms that the user holds the private key corresponding to the registered public key, and authentication is successful; otherwise, verification fails, and the authentication request is rejected. The entire process achieves strong authentication while maintaining complete confidentiality of the private key, ensuring that sensitive information is not leaked even if the communication channel is intercepted.

[0164] By implementing a zero-knowledge proof verification mechanism based on the SM3 hash algorithm, the system successfully achieved highly secure identity authentication. The main advantages of this method are: the private key remains confidential at all times and is not exposed even during the authentication process; it prevents man-in-the-middle and replay attacks by using a different random value for each authentication; it complies with domestic cryptographic regulations and adopts national standard cryptographic algorithms; it has low computational and communication overhead, making it suitable for implementation on various devices; and it provides mathematically provable security guarantees based on mature cryptographic theory. The zero-knowledge proof verification mechanism offers a secure yet privacy-conscious identity authentication method, making it particularly suitable for applications with high requirements for privacy protection and security.

[0165] Example 8:

[0166] In this embodiment, the deployment of an adaptive key update strategy, acquiring network threat situation data, analyzing threat level changes through a threat situation assessment function, and obtaining dynamically adjusted key update parameters includes: calculating a quantified threat level score based on the authentication results, the number of abnormal network logins, and the attack frequency using the threat situation assessment function, and generating a multi-dimensional threat feature vector; based on the multi-dimensional threat feature vector and historical threat data, performing intensive sampling in areas where the threat level score exceeds a preset threshold using an adaptive grid sampling strategy, and using a change point detection algorithm to identify key node information of threat situation changes, determining the triggering time for key updates; and based on the threat level score and the key node information, predicting threat development trends through a multi-level threat threshold system and a local function approximation method, and dynamically adjusting the key update frequency and strength parameters.

[0167] Specifically, the adaptive key update strategy is a dynamic security management mechanism that automatically adjusts the frequency and method of key updates based on the current threat landscape of the network environment. Unlike traditional fixed-period key update methods, the adaptive strategy analyzes the network security situation in real time, proactively strengthening key protection when high risks are detected and appropriately relaxing the update frequency during low-risk periods, achieving a balance between security and system resource consumption. This strategy is particularly suitable for identity authentication systems facing ever-changing security threats, flexibly responding to various threat scenarios ranging from low-intensity daily network noise to high-intensity targeted attacks. The basic principle of the adaptive strategy is "the higher the threat level, the more frequent and stronger the key updates." Through this dynamic response mechanism, the system can optimize resource utilization efficiency while maintaining high security.

[0168] The threat posture assessment function is a mathematical model that quantifies the current network security status, converting various security indicators into a unified threat level score. This function comprehensively considers multiple factors, including but not limited to: authentication result analysis, such as authentication failure rate and abnormal authentication patterns; abnormal network login behavior, such as login attempts at unusual times or locations; attack signature identification, such as detection results of brute-force attacks, dictionary attacks, or replay attacks; and abnormal system indicators, such as abnormal resource usage or log anomalies. The threat posture assessment function typically employs a weighted summation model, assigning different weights to different indicators based on their security importance. For example, multiple consecutive authentication failures may be given a higher weight, while occasional network latency may have a lower weight. The output of this function is a standardized threat level score, typically between 0 and 1, where 0 represents no threat and 1 represents an extremely high threat. The score directly influences subsequent key update decisions.

[0169] A multidimensional threat feature vector is a mathematical representation of the threat situation, organizing various security indicators into a structured vector form. Each dimension of this vector corresponds to a type of threat indicator, collectively describing the overall security situation faced by the system. Typical threat feature dimensions include: authentication anomaly indicators, such as failure authentication rate, number of consecutive failures, and failure distribution patterns; network behavior indicators, such as traffic anomalies, access frequency, and source IP diversity; time pattern indicators, such as attack duration, periodicity, and time distribution anomalies; and user behavior indicators, such as abnormal operation sequences, abnormal sessions, and abnormal permission usage. The system extracts and normalizes these features to form a standardized threat feature vector. This vector representation facilitates subsequent pattern recognition, cluster analysis, and anomaly detection, serving as the foundational data structure for threat situation assessment and prediction.

[0170] Historical threat data comprises long-term accumulated security event records and threat assessment results, providing a benchmark and trend basis for current threat analysis. Historical data typically includes the following: time series of past threat feature vectors; historical security event records, including attack type, intensity, and impact; past key update records and their effectiveness assessments; and confirmed attack patterns and signatures. The system maintains a structured historical threat database, ensuring the quality and usability of historical data through preprocessing steps such as data cleaning, noise reduction, and feature extraction. Timeliness management of historical data is also crucial; recent data is usually given higher weight, while older data may be downgraded or archived. By comparing with historical threat data, the system can identify emerging threat patterns, discover trends in attack intensity, and provide more comprehensive contextual information for current threat assessments.

[0171] Adaptive grid sampling is an intelligent data acquisition method that dynamically adjusts sampling density based on data characteristics to acquire more information in critical areas. In threat monitoring, this strategy optimizes the allocation of computing resources, focusing on high-risk areas. The implementation steps include: initializing a uniform grid to establish basic sampling points across the entire monitoring space; evaluating initial data to identify high-risk areas with threat level scores exceeding a preset threshold; refining the grid by increasing sampling point density in high-risk areas to form locally finer grids; iterative optimization by dynamically adjusting the grid density based on the evaluation results of new sampling points; and convergence judgment by completing the current sampling cycle when the rate of change in threat assessment of sampling points falls below a threshold. The core advantage of adaptive grids lies in their efficiency: maintaining sparse sampling in low-risk areas conserves resources; and performing dense sampling in high-risk areas ensures the capture of critical information. This "combination of extensive and intensive sampling, with a focus on key areas" approach enables the system to acquire the most valuable threat information with limited resources.

[0172] Change point detection algorithms are statistical methods for identifying key change points in time series data to detect significant shifts in threat posture. In the field of cybersecurity, change points typically represent a sudden increase in threat level or a change in attack patterns, and are important triggering factors for key update decisions. Commonly used change point detection algorithms include: CUSUM algorithm, which monitors the cumulative changes in data deviating from the expected average; Bayesian change point detection, which identifies changes in probability distribution based on a Bayesian statistical model; window comparison method, which compares the differences in statistical characteristics within adjacent time windows; and wavelet analysis, which uses wavelet transform to detect signal changes at different time scales. The implementation process of change point detection includes: data preprocessing, smoothing noise and standardizing data; algorithm selection, selecting an appropriate detection algorithm based on the characteristics of the threat data; parameter adjustment, setting detection sensitivity and false positive tolerance; multi-scale analysis, performing detection at different time granularities; and result verification, confirming the reliability of the detection results through cross-validation.

[0173] Key node information refers to the time points and related characteristics of significant changes in the threat landscape identified by the change detection algorithm. Each key node records the following information: timestamp, precisely identifying the time of the change; magnitude of change, quantifying the degree of increase or decrease in threat level; rate of change, describing the speed at which the threat landscape changes; persistence of change, assessing whether the change is a temporary fluctuation or a long-term trend; and related threat characteristics, identifying the main threat factors leading to the change. The system classifies and prioritizes key nodes, distinguishing between different types of threat changes: gradual change, where the threat level rises slowly but steadily; sudden change, where the threat level rises significantly in a short period; periodic change, where the threat level fluctuates regularly; and abnormal fluctuation, where unexpected changes do not conform to historical patterns. Key node information provides precise triggering criteria for key update decisions, enabling the system to initiate the update process at the most appropriate time.

[0174] The timing of key updates is the decision point by which the system decides when to initiate the key update process, directly impacting system security and resource efficiency. Based on key node information, the system employs a multi-factor decision-making mechanism to determine the triggering timing: absolute threshold triggering, triggered immediately when the threat level score exceeds a predetermined threshold; rate of change triggering, triggered when the threat level increases at a rate exceeding a threshold; persistence triggering, triggered after a threat level above the baseline has persisted for a preset time; pattern matching triggering, triggered when the current threat pattern matches a known high-risk attack pattern; and predictive triggering, triggered in advance based on threat trend predictions. The triggering decision also considers system status and business needs, such as: avoiding peak business periods to reduce service impact; considering system resource conditions to prevent large-scale updates during resource-constrained periods; and balancing update frequency to avoid excessive system overhead due to overly frequent updates. This multi-dimensional, context-aware triggering mechanism ensures that key updates occur when most needed, guaranteeing security while optimizing resource utilization.

[0175] A multi-level threat threshold system is a hierarchical threat assessment framework that sets differentiated trigger conditions for different levels of security response. This system typically includes 3-5 threat levels, each corresponding to a different key update strategy. A typical multi-level system includes: Normal level (0-0.2), maintaining a regular update cycle, such as once every 30 days; Warning level (0.2-0.4), moderately shortening the update cycle, such as once every 15 days, while increasing monitoring frequency; Alert level (0.4-0.6), significantly shortening the update cycle, such as once every 7 days, and strengthening key strength; High-risk level (0.6-0.8), adopting a short-term update strategy, such as once every 3 days, while activating emergency response plans; and Emergency level (0.8-1.0), triggering immediate updates, using the highest strength key parameters, and activating comprehensive protection measures. The multi-level thresholds not only consider the current threat score but also combine the rate of change and duration to form a dynamically adjustable decision boundary. This hierarchical response mechanism enables the system to take appropriate key protection measures according to the severity of the threat, achieving optimal resource allocation.

[0176] Local function approximation is a mathematical technique used to predict the behavior of a function in unsampled regions based on a finite number of data points. In threat trend prediction, this method can infer potential changes in threat levels in the short term based on observed threat data. Common local approximation techniques include: polynomial fitting, which uses low-order polynomial functions to fit local data points; spline interpolation, which uses piecewise polynomials to construct a smoothly transitioning fitting curve; locally weighted regression, which assigns higher weights to recent data points for local fitting; and radial basis functions, which use basis functions centered on data points to construct an approximation model. The implementation process of local function approximation includes: determining the prediction window, typically recent threat data points; selecting an appropriate local region to balance fitting accuracy and generalization ability; applying the approximation algorithm to construct a local function model; extrapolating prediction, which predicts short-term threat trends based on the constructed model; and uncertainty assessment, which quantifies the reliability of the prediction results. This method is particularly suitable for capturing nonlinear changes and short-term fluctuations in threat levels, providing a forward-looking basis for key update decisions.

[0177] Key update frequency and strength parameters are two core security parameters that the system dynamically adjusts based on the threat landscape. The frequency parameter controls the key update interval; as the threat level increases, the system shortens the update interval and increases the update frequency, and vice versa. The strength parameter controls the security characteristics of the key, including: key length, which determines the key's resistance to brute-force attacks; entropy source quality, affecting the key's randomness and unpredictability; algorithm complexity, determining the computational strength of key generation and verification; and multi-factor requirements, determining whether additional authentication factors are needed in key generation. The dynamic adjustment of these two parameters is based on the following rules: threat level is directly proportional to update frequency (higher threat, higher frequency); the rate of threat change is directly proportional to the frequency adjustment magnitude (rapid change, larger adjustment); sustained high threats trigger strength parameter increases to ensure long-term security; predicting trends influences advance adjustments to address potential risks. The system maintains a parameter adjustment matrix, mapping different threat levels to specific frequency and strength configurations, achieving fine-grained security control.

[0178] By deploying an adaptive key update policy, the system achieves intelligent response and proactive defense against network threats. This dynamic adjustment mechanism offers multiple advantages: resource optimization, avoiding resource waste or insufficient protection that can result from fixed-period updates; enhanced security, providing stronger key protection during high-threat periods; business continuity, minimizing the impact on normal operations by selecting appropriate update timings; and attack adaptability, capable of coping with constantly changing attack patterns and intensities. The adaptive policy shifts the identity authentication system from passive to proactive defense, not only responding to existing threats but also predicting and proactively addressing potential risks, forming a closed-loop security management mechanism. This intelligent key lifecycle management is a key component in building a resilient network security architecture, providing a dynamically adaptable security foundation for the entire system.

[0179] Example 9:

[0180] In this embodiment, establishing a secure communication channel based on SM4 symmetric encryption includes: generating a random session key and securely distributing it to both communicating parties using IBE encryption technology, based on the dynamically adjusted key update parameters and the IBE mechanism, to establish a temporary symmetric encryption key; encrypting the data using the SM4 symmetric encryption algorithm based on the temporary symmetric encryption key and the key distribution data to be transmitted, to obtain encrypted data; and verifying data integrity by calculating a message authentication code and appending it to the encrypted data based on the encrypted data and the SM4-GCM authentication encryption mode, thereby preventing tampering attacks during transmission and ensuring data transmission security during key distribution and update.

[0181] Specifically, during the key distribution phase, the system uses a temporary session key to encrypt transmitted data using SM4, and the session key is securely distributed via the IBE mechanism. A key escrow and backup mechanism is implemented, storing key fragments across multiple secure nodes to prevent key loss due to single points of failure. A key lifecycle management process is established, including security controls for key generation, distribution, use, update, revocation, and destruction. The SM4-GCM authentication encryption mode ensures data confidentiality while verifying data integrity, preventing tampering attacks during transmission.

[0182] A secure communication channel is a dedicated channel in a system to ensure secure data transmission, providing security features such as confidentiality, integrity, authentication, and non-repudiation. In key management systems, secure communication channels are particularly important because they carry the transmission of the most sensitive key materials. Unlike ordinary communication channels, secure communication channels use cryptographic techniques to ensure that data can securely reach its destination even in untrusted network environments. Establishing such a channel requires addressing several security challenges: the key distribution problem, i.e., how to securely share the initial encryption key; man-in-the-middle attack protection, ensuring the authenticity of the identities of both communicating parties; data confidentiality assurance, preventing eavesdropping on transmitted content; data integrity verification, detecting tampering during transmission; and replay attack protection, preventing the retransmission of historical messages. This embodiment employs an SM4-based symmetric encryption scheme, combined with IBE identity encryption and authentication encryption technologies, to construct a highly secure communication channel.

[0183] Identity-Based Encryption (IBE) is a special public-key encryption technology. Its key feature is that a user's public key can be directly derived from their identity information (such as email address, IP address, or device ID), without relying on a complex Public Key Infrastructure (PKI). An IBE system comprises four core components: a setup algorithm to generate the system master key and public parameters; a key generation algorithm where a key generation center uses the master key and user identity to generate a user's private key; an encryption algorithm to encrypt the message using the user's identity and public parameters; and a decryption algorithm to decrypt the ciphertext using the user's private key. The working principle of IBE is as follows: a central authority holds the system master key; the sender only needs to know the recipient's identity to encrypt the message using the system's public parameters; the recipient needs to obtain the private key corresponding to their identity from the central authority to decrypt the message. IBE simplifies the key management process, making it particularly suitable for key distribution in distributed environments and solving the complexity of certificate management in traditional PKI systems.

[0184] A random session key is a temporary key specifically generated for a single communication session and is discarded after use, significantly improving communication security. The generation process of the session key must ensure sufficient randomness and entropy, typically including the following steps: collecting entropy sources, obtaining randomness from multiple sources such as hardware random number generators, system events, or network time; seed processing, mixing and expanding the collected entropy to form a high-quality random seed; key derivation, deriving the session key from the seed using a cryptographically secure pseudo-random number generator (CSPRNG); and key verification, checking whether the generated key meets the strength requirements to avoid weak keys. The advantages of a random session key are: even if the key for one session is cracked, it will not affect the security of other sessions; the key usage period is short, greatly increasing the difficulty for attackers to crack it; and no long-term storage is required, reducing the risk of key leakage. In this system, the session key length is determined by dynamically adjusted key update parameters, with longer key lengths used in high-threat environments.

[0185] Secure distribution using IBE encryption technology is the process of securely transmitting session keys to both communicating parties using an identity-based encryption mechanism. In this system, IBE addresses the "chicken and egg" dilemma of traditional key distribution: it allows secure transmission of symmetric keys without the need for a pre-established secure channel. The steps of IBE encrypted distribution include: identity verification, verifying the identity of the communicating parties (such as device ID or user account); parameter acquisition, obtaining the latest IBE public parameters from the system; key encapsulation, encrypting a randomly generated session key using the recipient's identity and the IBE public parameters; ciphertext transmission, sending the encrypted session key to the recipient; private key acquisition, the recipient obtaining the IBE private key corresponding to their identity from the key generation center (this step is usually completed during system initialization); and key decryption, the recipient using their IBE private key to decrypt and obtain the session key. IBE technology ensures that only a legitimate recipient can decrypt and obtain the session key, and even if the communication channel is intercepted, the session key will not be leaked.

[0186] A temporary symmetric encryption key is a shared key used only in a single communication session, serving as input to the SM4 symmetric encryption algorithm. Unlike long-term keys, temporary keys have the following characteristics: short lifespan, typically lasting only for one session or one key update operation; destruction after use, securely removed from memory without leaving a trace after the session ends; independence, using a completely new key for each communication, with no correlation between them; and high randomness, generated by a cryptographically secure random number generator with high entropy. The process of establishing a temporary key includes: session initialization, where the initiator generates a random session key; key distribution, securely transmitted to the communicating party via IBE; synchronization confirmation, where both parties verify key consistency; and status recording, recording the key's usage status and validity period. The advantage of using temporary keys is that even if the key for a particular communication is compromised, the security of past and future communications remains guaranteed, achieving "forward secrecy" and "backward secrecy" characteristics.

[0187] Key distribution data is sensitive information that needs to be securely transmitted during the key update process. It includes: newly generated user key materials; key usage control parameters, such as validity period and usage permissions; key metadata, such as key version and generation timestamp; associated authentication data; and key verification information used by the recipient to verify the integrity and authenticity of the key. Secure transmission of key distribution data is a critical link in the entire key management system; leakage can lead to serious security problems. Before transmission, the system preprocesses the key distribution data: format standardization to ensure data conforms to a predetermined format; compression to reduce transmission volume; integrity verification and calculation for subsequent verification; and priority marking to indicate the urgency of processing. Key distribution data typically employs a layered protection mechanism: the inner layer uses strong encryption algorithms to protect the core key materials, while the outer layer uses authentication encryption to protect the entire transmission packet, forming a nested protection structure of "keys within keys."

[0188] The SM4 symmetric encryption algorithm is a block cipher standard designed to replace the internationally used AES algorithm, providing equivalent security strength and performance. SM4 uses a 128-bit key and a 128-bit block size, designed for efficient hardware and software implementation. Key features include: a 32-round nonlinear iterative structure providing ample confusion and diffusion; fixed S-box substitution enhancing resistance to analysis; a complex round key expansion mechanism ensuring each bit of the key significantly influences the ciphertext; and support for multiple operating modes such as ECB, CBC, CTR, and GCM. The SM4 encryption process includes: key expansion, generating 32 round keys from the 128-bit master key; initial data processing, dividing and arranging plaintext blocks according to a defined method; round function iteration, applying 32 rounds of transformation operations; and outputting the final ciphertext. In this system, SM4 is used to protect the confidentiality of key distribution data, ensuring that even if the data is intercepted, attackers without the key cannot access the content.

[0189] Encryption is the process of converting plaintext data into ciphertext using the SM4 algorithm, ensuring that only the recipient with the correct key can access the original information. The encryption process includes: data preparation, padding the data to be encrypted to a multiple of 128 bits (e.g., using PKCS#7 padding); mode selection, choosing an appropriate block cipher mode based on security requirements (this system uses GCM mode); initialization vector generation, creating a random initialization vector to ensure that the same plaintext produces different ciphertexts after encryption; block processing, dividing the data into 128-bit blocks and applying the SM4 algorithm to each block; and linking, linking the encrypted blocks together according to the selected mode. Key security considerations for encryption include: avoiding weak initialization vectors; using different initialization vectors for each encryption; ensuring that the padding method does not leak information; and correctly implementing all steps of the selected mode. The encrypted data contains the ciphertext itself and necessary auxiliary information (such as the initialization vector), which will be further authenticated and encrypted.

[0190] SM4-GCM authentication encryption mode is a working mode of the SM4 algorithm that provides both data confidentiality and integrity protection. GCM (Galois / Counter Mode) combines the efficient encryption of Counter Mode (CTR) with the message authentication capabilities of Galois field multiplication, making it one of the most popular authentication encryption modes currently available. The working principle of SM4-GCM includes: Encryption: Generating a keystream using Counter Mode and XORing it with the plaintext to obtain ciphertext; Authentication: Calculating an authentication tag using Galois field multiplication, overlaying the ciphertext and Additional Authentication Data (AAD). The execution flow of SM4-GCM includes: Initialization: Setting the initial counter value and authentication key; Additional Data Processing: Processing data that requires authentication but not encryption; Data Encryption: Encrypting the plaintext using Counter Mode; Tag Calculation: Calculating the authentication tag based on all processed data; Result Assembly: Combining the ciphertext, tag, and necessary parameters into the final output. The advantages of SM4-GCM are: strong parallel processing capability, supporting high throughput; providing confidentiality and integrity in a single processing operation; supporting additional authentication data to protect metadata other than plaintext; and pre-computation optimization to support efficient implementation in fixed-key scenarios.

[0191] Message Authentication Codes (AACs) are used to verify that data has not been tampered with during transmission and to confirm the authenticity of the data source. In SM4-GCM mode, AACs are tags generated based on Galois field multiplication. The generation process of AACs includes: initializing the hash subkey and generating a dedicated authentication subkey using the encryption key; processing the appended authentication data and calculating the hash value of the AAD; processing the ciphertext data and calculating the hash value of the ciphertext; length processing, incorporating the length information of the AAD and the ciphertext into the calculation; and tag generation, ultimately generating a fixed-length (usually 128-bit) authentication tag. The AAC is appended to the encrypted data to form a complete protection packet. The receiver verifies the AAC before decryption; only after successful verification will decryption proceed. This "authenticate before decryption" approach prevents attacks based on ciphertext operations, such as oracle padding attacks. The security strength of the AAC depends on the tag length and algorithm security; the 128-bit tag used in this system provides a high level of security.

[0192] Tampering attacks refer to attackers modifying data content during data transmission, attempting to change the meaning of communication or causing the receiver to perform unexpected operations. The harm of tampering attacks is particularly evident during key updates: it may lead users to obtain incorrect keys; it may cause the system to accept forged key materials; and it may trigger the system into an unexpected state. The SM4-GCM authentication encryption mode prevents tampering attacks through the following mechanisms: data binding, where the message authentication code tightly binds the ciphertext and metadata, and any modification will result in verification failure; full coverage protection, where authentication tag calculation covers all sensitive data, leaving no blind spots; key dependency, where the authentication process depends on the key, preventing attackers from generating valid tags without knowing the key; and strict verification, where the receiver strictly performs tag verification, rejecting any messages that fail verification. In addition, extra protective measures are implemented: a sequence number mechanism to prevent message reordering; timestamp verification to identify expired messages; and context binding, incorporating communication context information into the authentication scope.

[0193] Data transmission security is the fundamental guarantee for the entire key distribution and update process, ensuring the secure delivery of sensitive key materials from source to destination. This embodiment achieves comprehensive transmission security through three layers of protection: the first layer, session key protection, uses a temporary symmetric key distributed securely by IBE to encrypt the data content; the second layer, SM4 encryption protection, performs strong encryption processing on all key distribution data; and the third layer, GCM authentication protection, provides integrity and authenticity verification for encrypted data. This multi-layered protection ensures that data transmission security has the following characteristics: end-to-end protection, data is protected throughout the entire process from generation to receipt; anti-eavesdropping, encryption ensures that unauthorized parties cannot access the content; anti-tampering, authentication mechanisms ensure data integrity; anti-replay, using sequence numbers and timestamps to prevent the replay of historical messages; and identity binding, keys are closely associated with user identities to ensure correct delivery. Data transmission security is a key component of the entire identity authentication system, providing a reliable key distribution foundation for upper-layer applications.

[0194] By establishing a secure communication channel based on SM4 symmetric encryption, the system achieves advanced security assurance during key distribution and update processes. This scheme combines the flexible key distribution of IBE, the efficient encryption of SM4, and the authentication protection capabilities of GCM to form a complete secure communication solution. This secure communication channel is not only suitable for key distribution but can also be extended to other sensitive data transmission scenarios, providing a unified secure communication infrastructure for the entire identity authentication system. By adopting the national cryptographic standard algorithm SM4, it meets the requirements of national cryptographic regulations for commercial cryptographic applications while maintaining consistency with advanced international security practices, ensuring the system's compliance and security.

[0195] Example 10:

[0196] In this embodiment, based on the decentralized key distribution mechanism, a consortium blockchain network is constructed, and the key generation and distribution process is recorded on the blockchain, generating a key management blockchain record containing timestamps and digital signatures to ensure the transparency and immutability of the key management process. For the blockchain key management record, smart contracts are designed to automate key lifecycle management, including periodic updates, permission changes, and emergency revocation operations, generating smart contract execution logs and state change records. Based on the smart contract execution logs, the state change records, and the multi-signature mechanism, multiple authorized nodes are required to participate in critical key operations. The blockchain consensus mechanism is used to verify the legality of key operations, establishing a distributed trust system and enhancing the security of the decentralized key distribution mechanism.

[0197] Specifically, a decentralized key escrow solution based on blockchain constructs a consortium blockchain network, recording the key generation and distribution process on the blockchain to ensure the transparency and immutability of key management. Smart contracts are designed to automate key lifecycle management, including periodic updates, permission changes, and emergency revocation. A multi-signature mechanism requires multiple authorized nodes to participate in critical key operations, preventing malicious behavior by a single node. The consensus mechanism of the blockchain is used to verify the legitimacy of key operations, establishing a distributed trust system.

[0198] Decentralized key distribution (DHD) is a key management method that does not rely on a single central authority. It achieves key generation, distribution, and management through a distributed system. Compared to traditional centralized models, decentralized mechanisms eliminate the risk of single points of failure, improving system availability and resistance to attacks. The core concept of this mechanism is to distribute key management authority among multiple participants, ensuring that no single entity can completely control the key system. DHD typically employs threshold cryptography techniques, such as Shamir secret sharing or threshold signatures, to divide sensitive key materials into multiple shares held by different entities. Only when a sufficient number of shares are merged can the complete key be reconstructed or used. In practical deployments, DHD needs to address a series of challenges: initial trust establishment, determining the qualifications and permissions of the initial participants; share update mechanisms, updating the key shares held by each party without revealing the complete key; and consensus achievement, reaching agreement on key operations in a distributed environment. Blockchain technology provides an ideal platform for solving these challenges, and combined with smart contracts, it can achieve secure and transparent decentralized key management.

[0199] A consortium blockchain network is a semi-decentralized distributed ledger system maintained by multiple pre-selected organizations. Unlike public blockchains, consortium blockchains have access controls for participating nodes; only authorized institutions can join the network and participate in the consensus process. This design retains the decentralized nature of blockchain while providing higher performance and privacy protection. In key management scenarios, consortium blockchain networks typically consist of multiple trusted institutions, such as authentication service providers, enterprise security departments, and regulatory agencies. The steps to build a consortium blockchain network include: identifying participating organizations and selecting institutions with appropriate qualifications and trustworthiness; establishing a governance framework, clarifying the rights and responsibilities of each party and network operation rules; deploying infrastructure, building blockchain nodes and network connections; initializing the network, creating the genesis block and initiating the consensus mechanism; and integrating a security module and cryptographic hardware to enhance key protection. The core advantage of consortium blockchains lies in their balance between decentralization and performance requirements, while providing the privacy protection and access control functions required for enterprise-level applications.

[0200] A blockchain key management record is a permanent record on the blockchain of operations and state information at each stage of a key's lifecycle. These records include, but are not limited to: key generation events, recording the key's creation time, type, and purpose; key distribution operations, recording which entities the key was distributed to and when; key usage, recording the frequency and purpose of key use; key update events, recording the key update time and reason; and key revocation operations, recording the time and reason for key revocation. Each record is accompanied by a timestamp and a digital signature. The timestamp provides the precise time the operation occurred, while the digital signature verifies the record's authenticity and the operator's identity. Digital signatures are typically generated using asymmetric encryption technology. The operator uses their private key to sign the record content, and anyone can verify the signature's validity using the corresponding public key. The generation process of a blockchain record includes: constructing transaction data, including the operation type, key identifier, and related parameters; signing the transaction, using the operator's private key to sign the transaction data; submitting the transaction, sending the signed transaction to the blockchain network; consensus verification, where network nodes verify the transaction's validity and reach a consensus; and on-chain storage, packaging the verified transactions into blocks and storing them permanently.

[0201] A timestamp is a precise record of the time when a key operation occurs, and it is a crucial component of the blockchain record for key management. Timestamps are of particular importance in blockchain systems because they establish the temporal order of operations, helping to address issues such as double-spending. Blockchain timestamps typically consist of two levels: block timestamps, recording the time a block is generated; and transaction timestamps, recording the time a transaction is submitted to the network. Timestamps can be generated in several ways: system time synchronized via the Network Time Protocol (NTP); multi-node time consensus, taking the median of timestamps from multiple nodes; and trusted time services, connecting to an authoritative time service provider. To prevent timestamp tampering, multiple protection measures are employed: time window restrictions, rejecting timestamps that significantly deviate from network time; timestamp linking, ensuring each block's timestamp is later than the previous block; and consensus verification, where all nodes jointly verify the validity of the timestamps. Accurate and reliable timestamps are essential for key management, especially in scenarios such as determining key validity periods, identifying expired operations, and audit traceability.

[0202] A digital signature is an electronic signature generated using cryptographic techniques to verify the integrity of data and the identity of the sender. In a blockchain key management system, digital signatures ensure the authenticity and non-repudiation of each record. The generation process of a digital signature includes: generating a digest, calculating a digest of the data to be signed using a hash function; private key signing, encrypting the digest using the signer's private key to generate a signature value; and appending the signature value to the original data. The signature verification process includes: separating the data, separating the original data from the signature; calculating the digest, calculating a digest of the original data using the same hash function; public key verification, decrypting the signature value using the signer's public key and comparing it with the calculated digest; and result determination, if the two digests match, the signature is valid. In blockchain key management systems, digital signatures typically employ algorithms such as Elliptic Curve Digital Signature Algorithm (ECDSA) or SM2 signature algorithm, which provide high security and short signature lengths. To enhance security, the system usually stores the signing private key in a Hardware Security Module (HSM) to prevent the signing key from being leaked or misused.

[0203] Smart contracts are automatically executing program code deployed on a blockchain, capable of processing data and triggering operations according to predetermined rules, without third-party intervention. In key lifecycle management, smart contracts automate various key operations, ensuring these operations follow predefined rules and processes. The main functions of key management smart contracts include: key registration, recording metadata and permission information for new keys; periodic updates, automatically initiating key update processes based on preset time intervals or trigger conditions; permission management, controlling which entities can access or manipulate keys; state transitions, managing various state changes of keys from generation and activation to revocation; and emergency response, automatically executing key revocation operations when a security threat is detected. Smart contract design follows these principles: the principle of least privilege, ensuring contracts possess only the minimum permissions required to perform their functions; modular design, separating different functions into independent modules for easy upgrades and maintenance; fault tolerance, defaulting to the safest action in abnormal situations; and auditability, recording all operations and state changes for subsequent auditing. Before deployment, contracts undergo rigorous security audits and formal verification to ensure the code is free of vulnerabilities or logical errors.

[0204] Key lifecycle management encompasses the entire process from key creation to destruction, ensuring proper protection and use of keys throughout their lifetime. A complete key lifecycle includes the following stages: key generation (creating new keys that meet security requirements); key distribution (securely delivering keys to authorized users); key storage (securely storing key materials); key usage (using keys in authorized operations); key update (updating or rotating keys at appropriate times); key revocation (stopping the use of keys when they are leaked or no longer needed); and key destruction (permanently deleting key materials). Smart contracts automate the management of these lifecycle stages, specifically through: a periodic update mechanism that automatically adjusts the key update frequency based on risk assessment results; event-triggered updates that immediately initiate key updates when a preset security event is detected, including but not limited to: abnormal key usage frequency, access attempts from uncommon geographical locations, multiple consecutive authentication failures, system detection of intrusion, or receiving vulnerability warnings from authoritative institutions; permission change management that handles key access control adjustments caused by changes in user roles or permissions; and an emergency revocation process that responds quickly to key leaks or other security events, immediately revoking affected keys. Smart contracts ensure these operations are executed according to predefined rules, reducing human error and improving response speed.

[0205] A smart contract execution log is a detailed record of all operations and events during the execution of a smart contract, providing a complete audit trail of contract behavior. Execution logs typically include the following information: operation time, recording the precise timestamp of each operation; operation type, identifying the specific function performed, such as key generation, update, or revocation; input parameters, recording the parameter values ​​provided to the contract; execution result, recording the success or failure status of the operation; event notifications, recording event notifications triggered by the contract; and resource consumption, recording the computational resources consumed by the operation. The generation mechanism of smart contract execution logs includes: built-in log functions, embedding dedicated log recording statements in the contract code; event triggers, defining and triggering event notifications predefined by the contract logic; and state recording, automatically recording significant changes in the contract state. Execution logs are not only stored on the blockchain to ensure their immutability but may also be synchronized to external auditing systems for more complex analysis and monitoring. Comprehensive execution logs are crucial for troubleshooting, security auditing, and compliance verification, providing full visibility and traceability of key operations.

[0206] State change logs are specialized records that track changes in the state of keys and their associated entities, reflecting key transitions throughout the key's lifecycle. A state change log typically includes the following elements: key identifier, uniquely identifying the affected key; previous state, the state value before the change; subsequent state, the state value after the change; change time, the precise time of the state change; triggering reason, the event or condition that caused the state change; authorization information, the entity that approved the state change; and associated operations, other operations related to the state change. A key may have multiple states: generating, the key is being created; active, the key is usable; paused, the key is temporarily unavailable; about to expire, the key is nearing its scheduled expiration time; revoked, the key has been explicitly revoked; archived, the key is no longer active but the record remains; and destroyed, the key material has been permanently deleted. State change logs are stored on the blockchain in the form of transactions, ensuring their immutability and traceability. These records are crucial for the auditing and compliance of key management systems, answering key questions such as "when, who, and why," and providing necessary evidence for security incident investigations and compliance verification.

[0207] Multi-signature is a cryptographic technique that requires multiple private keys to authorize an operation. In blockchain key management, it enhances security and implements checks and balances. The basic principle of multi-signature is to set a threshold m-of-n, meaning that at least m signatures are required from n authorizers to approve an operation. Typical configurations include 2-of-3 (two signatures from three authorizers) or 3-of-5 (three signatures from five authorizers). The implementation process of multi-signature includes: signature request (initiating a request for a key key operation); notification distribution (the system notifies all authorized nodes of the operation pending approval); independent signing (each authorized node independently verifies the legality of the operation and decides whether to sign); signature collection (the system collects signatures from each node); threshold verification (checking if the collected signatures reach a predetermined threshold); and operation execution (when the number of signatures reaches the threshold, the requested operation is executed). Multi-signature is suitable for various key operation scenarios: key generation (ensuring the creation of a new key is recognized by multiple parties); key revocation (preventing a single entity from revoking a key without authorization); permission change (controlling modifications to key access permissions); and system parameter adjustment (such as changing security policies or threshold settings). Multi-signature effectively prevents internal threats and single points of failure.

[0208] Authorized nodes are member nodes in a consortium blockchain network that are granted specific permissions related to key management and network governance, enabling them to participate in key management decisions and operational verification. Authorized nodes are typically operated by different stakeholders, such as identity providers, security service providers, enterprise user representatives, or regulatory bodies. The setup of authorized nodes follows these principles: separation of responsibilities (different types of operations are approved by different combinations of authorized nodes); balance of interests (ensuring representation of different stakeholders); geographical distribution (nodes are distributed across different physical locations to enhance system availability and resilience); and security level (nodes must meet minimum security standards and compliance requirements). The responsibilities of authorized nodes include: participating in the consensus process and verifying the validity of blocks and transactions; signing critical operations as part of a multi-signature mechanism; maintaining network security, monitoring and reporting suspicious activity; and implementing governance decisions, participating in the adjustment of network rules and parameters. The identity and permission information of authorized nodes are stored in special smart contracts on the blockchain. The addition or removal of any node requires confirmation through multi-party consensus, ensuring the fairness and security of the authorization mechanism.

[0209] A blockchain consensus mechanism is an algorithm and protocol used by nodes in a distributed network to reach agreement on transaction order and system state. In a key management system, the consensus mechanism ensures that all key operations are correctly verified and recorded by the network. Consortium blockchains typically employ the following consensus algorithms: Practical Byzantine Fault Tolerance (PBFT), which provides high performance and instant finality, suitable for key operations requiring rapid confirmation; Proof-of-Authority (PoA), where pre-selected authority nodes take turns generating blocks, balancing decentralization and efficiency; and the Raft consensus algorithm, which is simple and efficient but has lower tolerance for node failures. The main steps of the consensus process include: transaction submission, where operation requests are submitted to the network; transaction verification, where nodes verify the format, signature, and permissions of the operation; consensus reaching, where nodes reach an agreement on the validity of the operation using the selected consensus algorithm; block generation, where verified operations are packaged into a new block; and state update, where all nodes update their local state. The consensus mechanism plays a role in verifying the legitimacy of key operations by: ensuring that operations follow predetermined rules, such as permission checks and business logic verification; preventing double execution and avoiding the same key operation being processed repeatedly; maintaining the order of operations and ensuring that key operations are executed in the correct order; and rejecting illegal operations by identifying and rejecting requests that do not conform to the rules.

[0210] A distributed trust system is a trust model that does not rely on a single central authority. It establishes system credibility through multi-party collaboration and technical mechanisms. In key management scenarios, distributed trust systems solve the single point of trust problem inherent in traditional centralized models. The core elements of this system include: multi-party participation, with different stakeholders jointly maintaining the system; a consensus mechanism to ensure all parties agree on the system state; cryptographic verification, using mathematical methods to verify the correctness of operations; transparency, with all operations and rules visible to participants; immutable records, with historical operations permanently recorded and unalterable; and an incentive mechanism to encourage honest behavior and punish malicious behavior. The establishment process of a distributed trust system includes: initial trust establishment, confirming initial participants through offline authentication or existing trust relationships; rule formulation, clarifying the rules for system operation and the responsibilities of each party; technical implementation, deploying a technical framework supporting distributed trust; a governance mechanism, establishing mechanisms for resolving disputes and system evolution; and continuous verification, maintaining trust through regular audits and real-time monitoring. This system is particularly suitable for key management scenarios involving multiple organizations, ensuring that even if some participants fail or act maliciously, the system as a whole remains secure and trustworthy.

[0211] By implementing a decentralized key distribution mechanism and combining it with blockchain technology, a secure and transparent key management infrastructure has been successfully built. The consortium blockchain provides an ideal platform, recording all key lifecycle events and ensuring these records are immutable. Smart contracts automate complex key management processes, reducing human error and improving response speed. Multi-signature mechanisms and distributed consensus further enhance system security, ensuring that critical operations are verified and authorized by multiple parties. This comprehensive solution not only addresses the single point of failure and internal threat issues faced by traditional centralized key management systems but also provides a complete audit trail, meeting the transparency and traceability requirements of enterprises and regulators. The combination of blockchain technology and key management represents an innovative direction in the security field, laying the technological foundation for building a more secure and reliable identity authentication system.

[0212] Example 11:

[0213] In this embodiment, based on the batch-updated user key and the NTRU lattice basis algorithm, a quantum-resistant hybrid key system is constructed by integrating lattice-based and encoding-based cryptographic algorithms to generate a quantum-secure user private key and a quantum threat assessment model. Based on quantum channel availability, traditional key management requirements, and the quantum-secure user private key, a quantum-secure key transmission mechanism is established by designing a fusion protocol of quantum key distribution and traditional key management, prioritizing quantum key distribution when the quantum channel is available. Based on the quantum threat assessment model, the quantum-secure key transmission mechanism, and the current cryptographic algorithm strength, the quantum security of the batch-updated user key is improved by dynamically adjusting the key length and encryption parameters of the subsequent quantum cryptographic algorithm. Here, the current cryptographic algorithm strength refers to the equivalent security strength of the currently used master cryptographic algorithm (such as SM2, NTRU, McEliece, etc.) under the current threat environment, typically measured by the computational effort required to break the algorithm (e.g., the number of bits equivalent to the symmetric key length). The methods for obtaining the current strength of cryptographic algorithms include: 1) querying and integrating the latest algorithm strength assessment reports and recommendations released by international and domestic cryptography standards organizations to obtain the standard strength value of each algorithm; 2) integrating a local or cloud-based threat intelligence analysis module, which continuously monitors the latest academic research progress in the field of cryptanalysis, discovered algorithm vulnerabilities, and potential new attack algorithms (especially progress against quantum computing attacks), and dynamically assesses the decay of the strength of each algorithm accordingly; 3) combining the standard strength and threat intelligence to generate a real-time strength score or level for each cryptographic algorithm at the current moment. For example, the current algorithm strength score of the SM2 algorithm may be lowered when quantum threats intensify. This real-time strength value is used as one of the input parameters for the quantum threat assessment model to decide whether to switch to a quantum-safe algorithm or increase the key length.

[0214] Specifically, the quantum-safe post-quantum cryptography enhancement scheme integrates lattice-based and encoding-based cryptographic algorithms to construct a hybrid key management system resistant to quantum attacks. The NTRU lattice basis encryption algorithm is used to generate quantum-safe user private keys, and the McEliece encoding cryptosystem is used to protect the key transmission process. A fusion protocol of quantum key distribution and classical key management is designed, prioritizing quantum key distribution when a quantum channel is available, otherwise reverting to the post-quantum cryptography scheme. A quantum threat assessment model is established to dynamically adjust the cryptographic algorithm strength and key length.

[0215] Batch-up user key updates refer to the generation and distribution of new key materials to multiple users simultaneously when preset conditions are triggered. Compared to single key updates, batch updates offer higher efficiency and resource utilization, making them particularly suitable for large-scale identity authentication systems. Batch updates are typically triggered by the following situations: periodic security policy requirements, such as quarterly key rotation; security threat detection, such as the discovery of potential vulnerabilities; algorithm upgrades, such as the need to migrate to a stronger encryption algorithm; and changes in user status, such as simultaneous adjustments to the permissions of a large number of users. The batch update process includes: batch planning, determining the update batch based on system load and priority; parallel generation, generating multiple keys simultaneously using multi-threaded or distributed computing; batch distribution, efficiently transmitting key materials through secure channels; synchronous activation, coordinating the effective time of new keys; and old key processing, securely destroying or archiving expired keys. In the context of quantum computing threats, batch updates also need to consider the quantum resistance of the key algorithm to ensure that the updated keys can resist quantum computing attacks. This necessitates the introduction of post-quantum cryptographic algorithms, such as lattice-based cryptographic algorithms like NTRU.

[0216] NTRU is a post-quantum cryptographic algorithm based on lattice cryptography, considered resistant to quantum computer attacks. NTRU's security is based on the Shortest Vector Problem (SVP) and Closest Vector Problem (CVP) in lattices, problems considered computationally difficult even for quantum computers. The basic principle of NTRU is to perform operations on a truncated polynomial ring, using two polynomials as the private key, and generating the public key through a convolution operation with modular reduction. The encryption process includes: selecting a random polynomial, performing polynomial operations with the public key and plaintext; and applying modular operations to generate ciphertext. The decryption process includes: performing inverse operations using the private key polynomial; and recovering the original plaintext through error correction steps in its standard decryption procedure. Compared to traditional public-key algorithms such as RSA, NTRU has the following advantages: higher computational efficiency, with faster key generation, encryption, and decryption operations; smaller key and ciphertext sizes, reducing communication and storage overhead; and quantum resistance, resisting quantum attacks such as Shor's algorithm. NTRU has undergone years of cryptanalysis and optimization, becoming an important candidate algorithm for post-quantum cryptography standardization.

[0217] Lattice-based cryptography is a cryptographic system built upon mathematical problems in lattice theory, and is considered resistant to quantum computing attacks. A lattice is a discrete group of adders in n-dimensional Euclidean space, generated by a set of linearly independent basis vectors. The security of lattice cryptography relies on two mathematical problems considered difficult even for quantum computers: the Shortest Vector Problem (SVP), which finds the shortest non-zero vector in the lattice; and the Closest Vector Problem (CVP), which finds the nearest lattice point to an out-of-the-lattice point. Major lattice-based cryptographic algorithms include: NTRU, a public-key encryption system based on polynomial rings; Ring-LWE, an encryption scheme based on learning errors on rings; FALCON, a digital signature algorithm based on the NTRU lattice; and Dilithium, a digital signature scheme based on learning errors on modular lattices. The implementation of lattice-based cryptographic algorithms typically includes the following steps: parameter generation, selecting appropriate dimensions and moduli according to security requirements; key generation, creating lattice-based public-private key pairs; encryption / signature, using the public key and random numbers; and decryption / verification, using the private key to recover the information or verify the signature. The main advantages of this type of algorithm are its high computational efficiency, moderate key size, and resistance to known quantum algorithm attacks, making it the main development direction of post-quantum cryptography.

[0218] Encoding-based cryptographic algorithms utilize challenges in coding theory to construct cryptographic systems, primarily based on the computational difficulty of decoding random linear codes. The core principle of encoding-based cryptographic algorithms is: transforming an easily decoded linear code (such as Goppa codes) into a seemingly random but difficult-to-decode linear code through a secret transformation; the sender uses a seemingly random public key to encode information and adds a finite number of errors; the receiver uses a private key to recover the original encoding and correct errors to obtain the original information. Major encoding-based cryptographic algorithms include: Classic McEliece, an improved version of the original McEliece system based on binary Goppa codes; BIKE (Bit Flipping Key Encapsulation), a key encapsulation mechanism based on quasi-cyclic codes; and HQC (Hamming Quasi-Cyclic), a scheme combining Hamming metrics and quasi-cyclic codes. The implementation steps typically include: generating an efficient and decodeable linear code; creating a mask transformation to hide the code structure; the public key contains the transformed generator matrix, and the private key contains the original code and transformation information; encryption is performed using the public key to encode and add an error vector; decryption uses the private key to recover the original encoding and correct errors. The greatest advantage of coding-based cryptographic algorithms lies in their long-term security research; no effective cracking methods, including quantum algorithm attacks, have been found in decades.

[0219] A quantum-resistant hybrid key system is a comprehensive security architecture built by combining multiple post-quantum cryptography techniques, designed to resist the threat of quantum computers. Compared to single-algorithm schemes, the hybrid system employs an "algorithm diversity" strategy, ensuring the overall system remains secure even if one algorithm is compromised. The design principles of the hybrid key system include: algorithmic complementarity (selecting algorithms based on different mathematical problems); security level matching (ensuring consistent security strength across components); performance balance (selecting suitable algorithm combinations for different scenarios); and graceful degradation (maintaining basic security even when some components are attacked). In this embodiment, the hybrid system primarily integrates a lattice-based NTRU algorithm and a selected encoding-based algorithm. The construction process includes: algorithm selection (choosing complementary algorithms based on security and performance requirements); parameter determination (setting algorithm parameters that meet security objectives); key generation (creating a composite key structure containing key materials for multiple algorithms); operational flow design (defining how to use various algorithm components in different scenarios); and key management mechanism (handling the storage, distribution, and updating of the composite key). The hybrid key system provides "defense depth" for the system, significantly improving its ability to resist quantum computing attacks; even if a certain algorithm is compromised, the system maintains overall security.

[0220] A quantum-safe user private key refers to an encryption key that can resist quantum computing attacks. Generated using post-quantum cryptography algorithms, its security strength is sufficient to cope with quantum computing threats. Unlike traditional private keys, quantum-safe private keys typically have different structures and larger sizes, reflecting the characteristics of post-quantum algorithms. In this embodiment, the quantum-safe private key is generated based on a hybrid key system, featuring the following characteristics: a two-layer structure, including lattice-based and encoding-based private key components; an enhanced entropy source, using a quantum random number generator or high-entropy source to provide true randomness; a larger key length to resist acceleration attacks by quantum algorithms; and a composite format, encapsulating private key materials from multiple algorithms using a standardized format. The private key generation process includes: parameter initialization, setting algorithm parameters according to security requirements; random seed generation, obtaining a high-quality random source; NTRU private key generation, creating a lattice-based private key component; encoding private key generation, creating an encoding-based private key component; private key combination, securely combining private key materials from different algorithms; and key protection, applying additional protection mechanisms such as encryption and segmented storage. Quantum-secure private keys are not only used for current encryption and authentication operations, but also provide long-term protection against potential future quantum computing threats, ensuring the forward security of the system.

[0221] The Quantum Threat Assessment Model is a systematic approach to analyze and quantify the potential threats posed by quantum computing to existing cryptographic systems. This model combines technology development prediction, algorithm complexity analysis, and a risk management framework to help organizations understand, predict, and respond to the security challenges posed by quantum computing. The core components of the Quantum Threat Assessment Model include: quantum capability prediction, estimating the timeline of quantum computer development and the growth curve of computing power; algorithm vulnerability analysis, assessing the sensitivity of various cryptographic algorithms to quantum attacks; asset valuation, identifying critical information assets requiring long-term protection; risk exposure time, analyzing the relationship between the duration of information confidentiality requirements and the development of quantum capabilities; and adaptive path planning, developing strategies for a gradual migration to quantum-safe algorithms. The model's construction steps include: threat intelligence gathering, tracking progress and breakthroughs in quantum computing research; algorithm analysis, assessing the complexity reduction of various cryptographic algorithms under quantum attacks; security parameter calculation, determining the required key length at different quantum threat levels; risk probability estimation, calculating the probability of the system being attacked by quantum attacks within a given timeframe; impact assessment, analyzing the potential consequences of a breach of the cryptographic system; and mitigation strategy development, designing and implementing corresponding protective measures. This model provides a structured approach to addressing the uncertainty of quantum threats, supporting evidence-based decision-making and resource allocation.

[0222] A quantum channel is a physical pathway for transmitting quantum information using the principles of quantum mechanics and is a core component of a quantum key distribution (QKD) system. Unlike classical communication channels, quantum channels transmit qubits, which are encoded in the physical properties of photons, such as polarization state, phase, or energy. The availability of a quantum channel refers to the system's ability to establish and maintain reliable quantum communication, and is affected by several factors: fiber quality (the degree of quantum signal loss in the transmission medium); distance limitations (quantum signal strength decays with distance, typically limited to 50-100 kilometers in QKD systems); environmental interference (the effects of temperature changes, vibration, and electromagnetic interference on quantum states); repeater technology (the availability and performance of quantum repeaters); and detector efficiency (the performance parameters of the photon detector at the receiving end). The assessment of quantum channel availability includes: link quality testing (verifying that the qubit error rate is within an acceptable range); key generation rate estimation (calculating the ratio of secure keys obtainable under given conditions); stability analysis (evaluating the system's performance under different environmental conditions); and percentage of available time (statistically calculating the proportion of time the channel can be used for key distribution). In hybrid systems, quantum channel availability is a key indicator for determining when to use quantum key distribution. The system continuously monitors this indicator and dynamically adjusts the key distribution strategy accordingly.

[0223] Quantum key distribution (QKD) is a technique that uses the principles of quantum mechanics to securely establish a shared key between two parties. Its security is based on the fundamental principles of quantum mechanics rather than computational complexity. The core advantage of QKD is its ability to detect any eavesdropping attempts because, according to the measurement principles of quantum mechanics, any measurement of a quantum state will cause a state change, which can be detected by the legitimate communicating party. Major QKD protocols include: the BB84 protocol, the first QKD protocol proposed by Bennett and Brassard, which uses the four polarization states of a single photon; the E91 protocol, a protocol based on quantum entanglement that uses Bell's inequality to detect eavesdropping; the BBM92 protocol, an entangled version of BB84; and the COW protocol, a coherent one-way protocol suitable for long-distance implementation. The basic workflow of a QKD system includes: quantum state preparation, where the sender generates a quantum state encoded with random bits; quantum transmission, where the quantum state is transmitted through a quantum channel; quantum measurement, where the receiver measures the received quantum state; basic alignment, where the communicating parties exchange partial measurement setup information through a public channel; error estimation, where the error rate of the qubits is evaluated to detect eavesdropping; and privacy amplification, where the final secure key is extracted through information-theoretically secure post-processing. The key generated by QKD possesses information-theoretically secure properties, guaranteeing security even against attackers with unlimited computing power—a characteristic that traditional cryptography cannot provide.

[0224] The fusion protocol of traditional key management and quantum key distribution is a hybrid architecture that combines the advantages of both technologies, providing a smooth transition and optimal security. The protocol is designed to consider various constraints in real-world deployment environments, such as the limited availability of quantum channels, the widespread presence of traditional systems, and compatibility requirements. The core mechanisms of the fusion protocol include: priority decision-making (QKD is prioritized when a quantum channel is available, otherwise it falls back to a post-quantum algorithm); key layering (QKD-generated keys are used to protect the most sensitive information or encrypt other keys); hybrid authentication (combining traditional digital certificates and quantum-secure authentication mechanisms); adaptive routing (dynamically selecting the key distribution method based on the availability of the quantum channel along the communication path); and seamless switching (automatically switching between quantum and traditional methods, transparent to upper-layer applications). The protocol implementation steps include: communication initialization (detecting quantum channel availability and negotiating the protocol version); channel evaluation (testing quantum channel quality and classical channel security); distribution strategy selection (determining whether to use QKD or a post-quantum algorithm based on the evaluation results); key negotiation (executing the key establishment process for the selected method); key verification (confirming that both parties have obtained the same key materials); and security parameter negotiation (setting parameters such as encryption algorithms and key update frequency). This fusion protocol achieves the principle of "quantum-safe as much as possible, and post-quantum-safe when necessary," maximizing the overall security of the system.

[0225] Quantum-secure key transmission mechanisms refer to key exchange and distribution systems capable of resisting quantum computing attacks. This embodiment combines quantum key distribution (QKD) with post-quantum cryptography algorithms. The mechanism employs a multi-layered protection strategy, adaptively selecting the optimal solution based on quantum channel availability. When the quantum channel is available, the system prioritizes quantum key distribution, providing information-theoretically secure key materials. When the quantum channel is unavailable, it falls back to key exchange based on post-quantum algorithms such as NTRU. The core components of the key transmission mechanism include: a channel monitor, which evaluates the quantum channel state and quality in real time; a transmission strategy engine, which determines the transmission method based on channel state and security requirements; a key encapsulation module, which securely encapsulates key materials using a selected method; a protocol conversion layer, which handles compatibility between different key protocols; and a key verification system, which ensures correct transmission of key materials. The specific transmission process includes: pre-transmission evaluation, detecting quantum channel availability and determining the strategy; key generation, creating the key materials to be transmitted; secure encapsulation, encrypting the key materials using a QKD-generated key or a post-quantum algorithm; transmission execution, sending the encrypted key materials through the selected channel; reception and verification, where the receiver decrypts and verifies the integrity and authenticity of the key; and confirmation and activation, where both parties confirm the key's availability and activate its use. This mechanism ensures secure key transmission under any network conditions, providing a solid security foundation for the identity authentication system.

[0226] Post-quantum cryptography is a cryptographic algorithm designed to resist quantum computing attacks. Its security is based on mathematical problems considered difficult even for quantum computers to solve. The main categories of post-quantum cryptography include: lattice cryptography, based on problems in lattices such as SVP and CVP; code-based cryptography, relying on the difficulty of decoding random linear codes; multivariate cryptography, based on the complexity of solving multivariate polynomial equations; hash-based cryptography, utilizing the one-way nature of hash functions; and homomorphic encryption, allowing computation on encrypted data without decryption. Each algorithm category differs in security parameters (such as key length), performance characteristics, and implementation complexity. Dynamically adjusting post-quantum algorithm parameters is an adaptive security mechanism that automatically optimizes the algorithm configuration based on current threat assessments and system requirements. The adjustment process includes: threat assessment, analyzing current quantum computing developments and cryptanalysis progress; security goal setting, determining the required security strength level; parameter calculation, calculating the required parameters using the latest cryptanalysis results; performance balancing, finding a balance between security requirements and system performance; and implementation updates, applying the adjusted parameters to the system. Commonly adjusted parameters include: key length, which directly affects the security strength of the algorithm; lattice dimension, which is crucial for lattice-based algorithms such as NTRU; error distribution parameter, which affects the security of LWE-based algorithms; and hash function selection, which has a significant impact on the security of hash-based signature schemes. Dynamic adjustment ensures that the system always uses the optimal parameter configuration that meets current security requirements.

[0227] By implementing this quantum-resistant security architecture, current and future quantum computing threats can be addressed. The hybrid key system combines the advantages of multiple post-quantum algorithms, providing algorithmic diversity and defensive depth. A quantum threat assessment model enables the system to predict and respond to security challenges arising from the development of quantum computing, achieving proactive protection. The integrated key transmission mechanism fully leverages the information-theoretic security properties of quantum key distribution while ensuring high security even when quantum channels are unavailable. A dynamic parameter adjustment mechanism allows the system to adapt to constantly changing threat environments, optimizing the balance between security and performance. This comprehensive quantum security strategy not only addresses current security needs but also provides a future-oriented technological path, ensuring a high level of security even in the era of quantum computing. By applying these advanced technologies, the identity authentication system can provide users with long-term security, effectively protecting user identity and data security even in a future environment where quantum computing capabilities have significantly increased.

[0228] Example 12:

[0229] In this embodiment, based on the threat situation analysis results of the dynamically adjusted key update parameters, distributed threat detection nodes are deployed and a threat identification model is collaboratively trained using federated learning technology to generate threat level assessment results and threat classification labels. For the threat level assessment results and threat classification labels, a multi-layered threat response mechanism is designed to automatically adjust key update policies, access control policies, and audit policies according to the threat level, generating threat response execution records and system security configuration adjustment logs. Based on the threat response execution records, the system security configuration adjustment logs, historical attack patterns, and the current system status, predictive security analysis is implemented and an artificial intelligence decision engine is integrated to predict potential security risks and deploy protective measures in advance.

[0230] Specifically, dynamically adjusted key update parameters are an adaptive security mechanism that adjusts the frequency, strength, and method of key updates in real time based on the current threat environment and system status. Unlike statically preset key update strategies, dynamic parameters can respond specifically to different security events, strengthening protection during high-threat periods and optimizing performance during low-threat periods. Key parameters include: update frequency, determining the time interval for key rotation; key strength, including key length and complexity; update scope, determining which users or system components need key updates; update priority, determining the update order of different key types; and update method, selecting the most suitable key distribution channel. Triggers for parameter adjustments include: security event detection, such as the discovery of intrusion attempts; threat intelligence updates, such as the discovery of new attack methods; changes in system load, leading to more comprehensive updates when resources permit; and abnormal user behavior, strengthening key protection for suspicious users. Threat situation analysis results are key inputs for evaluating these parameters, providing a comprehensive view of the current security environment, including attack types, strengths, sources, and potential targets, providing data support for key update decisions. Through this dynamic mechanism, the system can achieve an optimal balance between security and performance, avoiding resource waste and security vulnerabilities.

[0231] Threat posture analysis is a comprehensive method for assessing system security status. It constructs a holistic view of the current threat environment by collecting and analyzing multi-source security data. The core tasks of threat posture analysis are identifying potential threats, assessing their severity, and predicting possible attack paths. The analysis process includes: data collection, acquiring data from multiple sources such as network traffic, logs, and endpoint behavior; data standardization, converting data of different formats into a unified format; feature extraction, identifying key security indicators in the data; correlation analysis, establishing relationships between events and discovering attack chains; threat scoring, quantifying threats based on their severity, scope of impact, and urgency; and trend analysis, identifying attack patterns that change over time. The outputs of threat posture analysis include: current threat level, representing the overall threat level faced by the system; an active threat list, identifying currently active attacks; vulnerable assets, system components that may become targets; and recommended response measures, suggesting mitigation or defensive actions. In key management systems, threat posture analysis results directly influence the adjustment of key update parameters. For example, when a new attack targeting the encryption algorithm used by the system is detected, the system will increase the update priority of keys using that algorithm.

[0232] Distributed threat detection nodes are security monitoring components deployed in different locations across the network, collectively forming a comprehensive threat detection network. Compared to centralized monitoring, the distributed architecture offers broader coverage and greater resilience. Each node is an independent detection unit equipped with multiple security monitoring capabilities: network traffic analysis, monitoring network communication patterns and anomalies; behavioral analysis, monitoring user and system behavioral deviations; signature detection, identifying characteristic patterns of known threats; and anomaly detection, identifying activities deviating from the normal baseline. Node deployment strategies consider the following factors: network topology, ensuring coverage of critical network segments; asset value, deploying more nodes around high-value assets; threat exposure surface, strengthening monitoring of vulnerable areas; and performance impact, optimizing node configuration to minimize disruption to normal business operations. Each node has local analytical capabilities, enabling preliminary threat assessment, and then sending the results to a central analysis platform or sharing insights through federated learning without sharing raw data. Nodes communicate through secure channels, ensuring the integrity and confidentiality of threat intelligence. This distributed architecture improves the reliability and attack resistance of the detection system; even if some nodes are bypassed or disabled by attackers, the overall detection capability remains effective.

[0233] Federated learning is a distributed machine learning approach that enables multiple participants to collaboratively train a model without directly sharing raw data. In the security field, federated learning addresses data privacy and sensitivity issues, allowing different security nodes or organizations to jointly build more robust threat detection models while protecting their own data privacy. The basic principles of federated learning are: local training, where each participant trains the model using local data; model parameter sharing, where participants only share model parameters (such as gradients) rather than the raw data; parameter aggregation, where a central server or a decentralized approach aggregates the model parameters from all participants; and model updates, where the updated global model is distributed to all participants. In threat detection applications, the federated learning process includes: initialization, deploying a basic threat detection model to each node; feature engineering, where each node extracts security features from its local data; local training, where nodes optimize the model using local data; secure aggregation, where model updates are securely aggregated using cryptographic or differential privacy techniques; model evaluation, testing the detection performance of the updated model; and continuous improvement, where the model is continuously optimized as new data is generated. This collaborative approach enables the system to learn diverse attack patterns from different environments, significantly improving the accuracy and coverage of threat identification while avoiding the risks of centralized storage of sensitive security data.

[0234] Threat identification models are machine learning models specifically designed to detect and classify security threats. These models learn from a large number of features of normal and malicious behavior to identify and classify potential security threats. The model architecture typically combines multiple techniques: supervised learning, using labeled attack samples to train a classifier; unsupervised learning, detecting anomalous behavioral patterns; deep learning, using neural networks to capture complex attack features; and reinforcement learning, optimizing detection strategies to cope with constantly evolving attacks. The development process of a threat identification model includes: data collection, acquiring samples of various security incidents; feature extraction, extracting meaningful features from the raw data; label preparation, with security experts labeling the data; model selection, choosing an algorithm suitable for the characteristics of the problem; model training, training the model using the prepared data; performance evaluation, testing the model's performance on unseen data; and model deployment, applying the trained model to a real-world environment. Through federated learning, these models can learn from threat detection nodes distributed in different locations, gaining a more comprehensive threat awareness. The model's output includes threat level assessment results (indicating the severity of the threat) and threat classification labels (such as DDoS attacks, credential theft, zero-day vulnerabilities, etc.), which provide a basis for subsequent threat response decisions.

[0235] Threat level assessment results are a quantitative representation of the severity and urgency of identified security threats. Assessment results typically employ a multi-level classification system, such as a 5-level scale (Severe, High, Medium, Low, Informational) or a 10-level scale. Assessment dimensions include: potential impact (the extent of damage a threat may cause); attack complexity (the level of technical skill required to carry out an attack); exploitability (the probability of successfully exploiting a vulnerability); detection difficulty (the ease with which an attack can be detected); and response complexity (the resources and time required to mitigate the threat). The assessment process integrates multiple information sources: threat intelligence (the characteristics and severity of known attacks); asset value (the importance of affected systems); vulnerability data, such as CVSS scores; attack intent analysis (inferring targets based on attacker behavior); and historical impact (a record of similar attacks). Assessment results are dynamic and updated as threats evolve and more information becomes available. Higher threat levels typically trigger more urgent response measures, such as immediate key rotation, stricter access controls, or more frequent system monitoring. Threat level assessments provide a scientific basis for prioritizing security resource allocation, ensuring that systems address the most severe threats first.

[0236] Threat classification tags are standardized classifications of security threat types, helping organizations understand the nature of threats and select appropriate response strategies. Common classification criteria include: attack technique classification, such as the MITRE ATT&CK framework, which describes attacker tactics and techniques; threat source classification, such as insider threats, external attackers, and advanced persistent threats (APTs); attack target classification, such as data theft, service disruption, and ransomware; and attack phase classification, such as reconnaissance, initial access, and lateral movement. In this system, threat classification tags focus specifically on threat types related to key management: key disclosure attempts, attacks that attempt to obtain encryption keys; algorithmic attacks, attempts to crack encryption algorithms; man-in-the-middle attacks, intercepting or modifying key exchange communications; side-channel attacks, obtaining key information by analyzing system physical characteristics; and denial-of-service attacks, blocking the availability of key management services. The tag generation process includes: feature analysis, extracting features of security events; pattern matching, comparing features with known attack patterns; behavioral analysis, analyzing the intent and methods of attack behavior; expert rule application, applying classification rules defined by security experts; and confidence scoring, indicating the credibility of the classification results. Accurate threat classification is crucial for selecting effective response strategies. Different types of threats require different defense measures, especially in key management systems, where key update strategies may be completely different for different threat types.

[0237] A multi-layered threat response mechanism is a graded defense system that takes appropriate security measures based on the severity and type of the threat. This mechanism employs the "defense in depth" principle, providing comprehensive security protection through multiple layers of protection. Response layers typically include: an automated response layer (pre-configured responses to known threats); a tactical response layer (short-term targeted defensive measures); a strategic response layer (long-term security hardening and architecture adjustments); and a recovery response layer (system recovery and hardening after an incident). In a key management system, response measures at different levels include: early warning measures, such as increasing monitoring frequency or sending alerts; defensive measures, such as blocking suspicious IPs or strengthening authentication requirements; mitigation measures, such as emergency key rotation or session termination; and recovery measures, such as key reconstruction or system reset. The implementation process of the response mechanism includes: threat assessment (analyzing threat levels and classifications); response selection (selecting appropriate response measures based on threat characteristics); automated execution (automatically executing the selected responses); effect monitoring (evaluating the effectiveness of response measures); and situational adaptation (adjusting response strategies according to threat evolution). This multi-layered response mechanism ensures that the system can take the most appropriate action based on the specific circumstances of the threat, enabling both rapid response to urgent threats and systematic resolution of long-term security vulnerabilities.

[0238] Key update policies refer to how a system manages the lifecycle of keys, specifically when and how keys are replaced. In threat response, key update policies dynamically adjust the following parameters based on threat level: update frequency (shortening key lifespan during high-threat scenarios); update scope (identifying which users or system components require key updates); update priority (determining the update order for different key types); update method (selecting the most secure key distribution channel); and strength adjustment (using longer key lengths in high-threat environments). Typical adjustment logic includes: maintaining a normal update frequency for low-level threats (e.g., routine scans); increasing the key update frequency for affected components for medium-level threats (e.g., targeted intrusion attempts); performing emergency system-wide key rotation for high-level threats (e.g., zero-day vulnerabilities or insider threats); and prioritizing updates to keys using specific algorithms (e.g., new attacks targeting the encryption algorithms used by the system) and considering algorithm migration. The adjustment process is automatically executed by the policy engine, integrating threat analysis results and system resource status to ensure necessary security protection without excessive system resource consumption. This adaptive key update policy significantly improves the system's responsiveness to different threats and is a core component of dynamic security defense.

[0239] Access control policies define which users or systems can access which resources and under what conditions access is permitted. Dynamic adjustment of access control policies is a critical defense mechanism when responding to security threats. Adjustments include: permission levels (increasing authorization requirements for sensitive operations); authentication strength (adding additional authentication factors); session management (shortening session timeouts or limiting concurrent sessions); access path restrictions (restricting allowed network paths or devices); and monitoring strength (increasing the audit frequency for high-risk users). Policy adjustments for different threat levels include: for low-level threats, increasing monitoring of abnormal behavior; for medium-level threats, implementing stricter access time limits and source verification; for high-level threats, activating emergency mode, allowing only the minimum necessary access permissions; and for threats with clear attack characteristics (such as credential theft), temporarily disabling affected accounts and requiring re-authentication. Policy adjustments are executed through the access control engine and are tightly integrated with the identity management system and threat detection system. This dynamic access control mechanism can immediately reduce the system's attack surface when a threat emerges, while maintaining necessary access for legitimate users, achieving a balance between security and availability.

[0240] Audit policies define how the system records, monitors, and analyzes security-related activities. As threat levels change, audit policies need to be adjusted accordingly to provide more comprehensive visibility. Key adjustments include: audit scope (determining which types of events to log); audit detail (determining the amount of information to log for each event); audit frequency (setting the frequency of collecting and analyzing audit data); retention period (specifying the retention time for audit logs); and alert thresholds (setting the conditions for triggering security alerts). The logic for adjusting audit policies in threat response includes: for low-level threats, increasing the audit detail of relevant system components; for medium-level threats, expanding the audit scope and shortening the analysis cycle; for high-level threats, activating full audit mode to log all user and system activities; and for threats targeting high-value or critical infrastructure assets (such as root key generation centers and core databases), implementing special audit rules for these assets. Dynamic adjustments to audit policies not only help detect attack attempts promptly but also provide detailed evidence for security incident investigations. They automatically balance audit comprehensiveness with performance impact, ensuring necessary visibility in high-threat situations while avoiding excessive audit data generation under normal circumstances.

[0241] Threat response execution logs are detailed records of all actions taken by the system in response to a security threat, providing a complete audit trail for security operations and post-incident analysis. Execution logs typically include the following information: response time (the exact time the action was recorded); triggering conditions (the threat indicator or event that led to the response); response type (e.g., key update, access restriction, or audit enhancement); execution details (the specific operational parameters and scope); execution status (e.g., successful, partially successful, or failed); impact assessment (the impact of the response measures on the system and users); and authorization information (the entity that approved or executed the response). The log generation process includes: event capture (recording the triggering of the response decision); operation logging (detailing each step of the execution); result verification (confirming the effectiveness of the response measures); association tagging (associating the response with the original threat); and secure storage (protecting the log using tamper-proof mechanisms). The value of execution logs lies in several aspects: incident investigation (providing a complete response timeline); effect analysis (evaluating the effectiveness of different response strategies); compliance auditing (demonstrating the organization's appropriate response to security incidents); and improvement basis (providing data support for optimizing response mechanisms). The system ensures that all response actions are fully recorded, whether automatically executed or requiring manual intervention.

[0242] The system security configuration adjustment log is a dedicated log that records changes to system security parameters and settings. It complements the threat response execution log, together forming a complete security change history. The adjustment log records more granular system configuration changes, including: parameter name (the specific configuration item modified); original value (the configuration value before modification); new value (the configuration value after modification); change time (the exact time the configuration was changed); reason for change (the event or policy that triggered the configuration change); authorization information (the entity that approved the change); and scope (the affected system components). Configuration adjustment types include: algorithm parameters, such as encryption strength or hash iteration count; timeout settings, such as session or token expiration; threshold configurations, such as the number of failed attempts or rate limiting; filtering rules, such as IP blacklists or content filtering rules; and monitoring settings, such as log levels or alarm conditions. The management mechanisms for the configuration adjustment log include: change verification (confirming that configuration changes have been correctly applied); version control (maintaining historical versions of the configuration); rollback capability (restoring to a previous configuration when necessary); and difference analysis (comparing the configuration status at different points in time). These logs are crucial for understanding the evolution of the system's security posture, helping analysts track how system configurations respond to different threat scenarios and providing a reference for future security hardening.

[0243] Predictive security analytics uses advanced analytics to predict future security threats, enabling organizations to proactively address potential risks rather than reactively responding to attacks that have already occurred. This forward-looking approach integrates multiple technologies: time-series analysis to identify periodic patterns in security events; trend analysis to track the development of threat indicators; correlation analysis to discover relationships between different security events; pattern recognition to identify activities similar to known attack precursors; and anomaly detection to detect activities that deviate from normal behavior. The analysis process includes: historical data collection, integrating past security events and response records; feature engineering, extracting predictive indicators; model development, building mathematical models capable of predicting future threats; validation and testing, evaluating the model's predictive accuracy; and continuous learning, constantly optimizing the model as new data is generated. Key applications of predictive analytics include: attack path prediction, identifying the next steps attackers might take; vulnerability exploitation prediction, predicting which vulnerabilities are most likely to be exploited by attackers; resource allocation optimization, pre-deploying security resources in high-risk areas; and threat intelligence generation, creating customized threat indicators that reflect the organization's own risk characteristics and business environment. Through predictive security analytics, the system can identify and mitigate potential threats before attacks cause actual damage, fundamentally changing the passive nature of security defense.

[0244] An AI decision engine is a system that uses advanced AI technology to automatically assess security situations and make or recommend response decisions. Unlike traditional rule-based systems, AI decision engines can handle complex uncertainties and incomplete information, making more flexible and intelligent decisions. Core technical components include: machine learning models that learn effective response strategies from historical security events; reasoning systems that apply causal reasoning and logical analysis to make decisions; natural language processing that analyzes security reports and threat intelligence; knowledge graphs that represent relationships and dependencies between security concepts; and reinforcement learning that optimizes long-term security strategies through trial and error. The decision-making process includes: contextual understanding, integrating multi-dimensional information about the current security posture; risk assessment, evaluating the potential risks and benefits of different decision options; decision generation, generating possible response options and their expected effects; decision optimization, selecting the best response strategy; and execution supervision, monitoring decision execution and adjusting the response. The advantages of an AI decision engine lie in its ability to: process massive amounts of complex data, far exceeding human analytical capabilities; respond quickly, making decisions in the early stages of an attack; learn and adapt, continuously improving from new attack and defense experiences; and provide consistency, avoiding the inconsistencies of human judgment. AI decision engines are tightly integrated with other security components, enabling the entire security system to be adaptive and self-learning.

[0245] By implementing this comprehensive adaptive security system, this invention achieves intelligent detection, response, and prediction of threats. Distributed threat detection nodes and federated learning technology enable the system to comprehensively monitor the security posture while protecting the privacy of sensitive data. A multi-layered threat response mechanism ensures that the system can take the most appropriate defensive measures based on the severity and type of the threat, dynamically adjusting key management, access control, and auditing policies. Predictive security analytics and an AI decision engine further enhance the system's foresight, enabling it to anticipate and proactively respond to potential threats. This integrated approach not only improves the system's ability to resist current attacks but also establishes a continuously evolving security defense system capable of adapting to the ever-changing threat environment. Through detailed recording of threat response execution logs and configuration adjustment logs, the system also provides valuable data sources for security analysis and continuous improvement, forming a self-improving security closed loop. This intelligent, adaptive security architecture represents a new height in the defense capabilities of identity authentication systems, providing users with more reliable and secure authentication services.

[0246] Example 13:

[0247] like Figure 2 As shown, the present invention also provides an identity authentication system based on national cryptographic algorithms, comprising:

[0248] The hierarchical key management module 10 is used to establish a hierarchical key management system based on the national cryptographic IBE framework, obtain elliptic curve parameters and base point G, use the SM2 algorithm to generate system master key pairs and establish domain key generation centers for each identity domain, and realize a decentralized key distribution mechanism.

[0249] Privacy protection module 20 is used to construct a domain-aware differential privacy protection module based on the decentralized key distribution mechanism, obtain the security level information of the user identity domain, define a privacy budget allocation strategy according to the security level information, add calibrated Laplace noise to the user identity feature vector, and generate privacy-protected identity authentication data.

[0250] The batch key update module 30 is used to design a distributed batch matrix multiplication protocol based on the privacy-protected identity authentication data, obtain key update requests from multiple users, decompose them to multiple computing nodes for parallel processing through secret sharing technology, and generate batch-updated user keys.

[0251] The zero-knowledge verification module 40 is used to implement a zero-knowledge proof verification mechanism based on the batch-updated user key, and to construct a zero-knowledge proof protocol through a commitment scheme based on the SM3 hash algorithm, so as to complete the identity verification and generate the authentication result without exposing the user's private key.

[0252] The adaptive policy control module 50 is used to deploy an adaptive key update policy based on the authentication result, obtain network threat situation data, analyze the changes in threat level through a threat situation assessment function, and obtain dynamically adjusted key update parameters.

[0253] The secure communication module 60 is used to establish a secure communication channel based on SM4 symmetric encryption according to the dynamically adjusted key update parameters, and to use a temporary session key for encryption processing to ensure the security of data transmission during key distribution and update.

[0254] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An identity authentication method based on Chinese national cryptographic algorithms, characterized in that, Includes the following steps: Establish a hierarchical key management system based on the national cryptographic IBE framework, obtain elliptic curve parameters and base point G, use the SM2 algorithm to generate system master key pairs and establish domain key generation centers for each identity domain to realize a decentralized key distribution mechanism; Based on the decentralized key distribution mechanism, a domain-aware differential privacy protection module is constructed to obtain the security level information of the user identity domain, define a privacy budget allocation strategy according to the security level information, and add calibrated Laplace noise to the user identity feature vector to generate privacy-protected identity authentication data. Based on the privacy-protected identity authentication data, a distributed batch matrix multiplication protocol is designed to obtain key update requests from multiple users and decompose them into multiple computing nodes for parallel processing through secret sharing technology to generate batch-updated user keys. Based on the batch-updated user keys, a zero-knowledge proof verification mechanism is implemented. A zero-knowledge proof protocol is constructed using a commitment scheme based on the SM3 hash algorithm to complete identity verification and generate authentication results without exposing the user's private key. Based on the authentication results, an adaptive key update strategy is deployed, network threat situation data is obtained, and the threat level changes are analyzed through a threat situation assessment function to obtain dynamically adjusted key update parameters. Based on the dynamically adjusted key update parameters, a secure communication channel based on SM4 symmetric encryption is established, and a temporary session key is used for encryption to ensure the security of data transmission during key distribution and update.

2. The method according to claim 1, characterized in that, The establishment of a hierarchical key management system based on the national cryptographic IBE framework includes: Based on the SM2 elliptic curve algorithm and preset curve parameters, the system master key pair is generated through the root key generation center, and a secure elliptic curve base point is selected to obtain the system-level master private key and master public key. Based on the system-level master private key and the domain identifiers of each identity domain, a dedicated domain master key is calculated for each identity domain through key derivation functions and SM3 hash operations, and multiple domain key generation centers are established. Based on the user's identity identifier and the domain master key, the user's identity is mapped to a point on an elliptic curve through a double hash structure and scalar operations are performed in conjunction with the domain master key to generate a user-specific private key and realize the decentralized key distribution mechanism.

3. The method according to claim 2, characterized in that, The calculation of a dedicated domain master key for each identity domain using a key derivation function and SM3 hash operation includes: Based on the system-level master private key and the domain identifiers of each identity domain, the domain identifiers are hashed using a hash function based on SM3 to obtain the domain feature value, thereby generating a unique identifier parameter for the domain. Perform elliptic curve operations on the domain feature value and the system master private key to generate a domain master private key specific to each identity domain; Based on the domain master private key and elliptic curve basis points, the corresponding domain master public key is calculated through elliptic curve dot product operations, thereby establishing a domain-level key management system that supports the decentralized key distribution mechanism.

4. The method according to claim 1, characterized in that, The constructed domain-aware differential privacy protection module acquires the security level information of the user identity domain, defines a privacy budget allocation strategy based on the security level information, and adds calibrated Laplacian noise to the user identity feature vector to generate privacy-protected identity authentication data, including: Based on the decentralized key distribution mechanism and the data sensitivity of each identity domain, the privacy requirements of information within the domain are analyzed through a security level assessment algorithm to determine a tiered privacy budget allocation strategy. Based on the privacy budget allocation strategy and user identity feature vector, identity feature data with added privacy protection noise is obtained by calculating function sensitivity and generating calibrated Laplace noise; Based on the identity feature data with added privacy-preserving noise, a heterogeneous joint learning framework is used to collaboratively train an identity recognition model across different domains. Gradient compression and a secure aggregation protocol are employed to protect the model parameter exchange process, thereby obtaining the privacy-preserving identity feature data.

5. The method according to claim 4, characterized in that, The process of calculating functional sensitivity and generating calibrated Laplace noise includes: Based on the privacy budget allocation strategy and the various dimensions of the user identity feature vector, the function sensitivity values ​​of various identity features are determined through theoretical analysis and experimental verification, and the noise scale calculation parameters are obtained. Based on the noise scale calculation parameters and the privacy budget parameters in the privacy budget allocation strategy, calibrated Laplace noise is generated according to the ratio of the noise scale calculation parameters to the privacy budget parameters. The calibrated Laplace noise is added to each dimension of the user identity feature vector by vector addition, resulting in the identity feature data with added privacy protection noise.

6. The method according to claim 1, characterized in that, The designed distributed batch matrix multiplication protocol obtains key update requests from multiple users, decomposes them into multiple computing nodes for parallel processing using secret sharing technology, and generates batch-updated user keys, including: Based on the privacy-protected identity authentication data and key update requests from multiple users, serial key updates are transformed into batch matrix operations by constructing a user identity matrix and a key generation matrix, resulting in a matrix-based key generation task. Based on the matrixed key generation task, the user identity matrix and the key generation matrix are decomposed into multiple matrix shards and distributed to different computing nodes through the Shamir secret sharing scheme to achieve distributed storage of sensitive information. Based on matrix sharding of each computing node, the multiplication results of each shard are calculated by performing local matrix multiplication operations in parallel, and the final result is reconstructed using Lagrange interpolation to generate the batch-updated user key.

7. The method according to claim 1, characterized in that, The zero-knowledge proof verification mechanism is implemented by constructing a zero-knowledge proof protocol based on a commitment scheme using the SM3 hash algorithm. This completes authentication and generates authentication results without exposing the user's private key, including: Based on the batch-updated user keys and preset elliptic curve parameters, the corresponding identity public key is generated through discrete logarithm operation and stored in the authentication server, establishing a public-private key correspondence. Based on the randomly selected temporary value, the elliptic curve parameters, and the user's private key in the public-private key correspondence, a response value containing the user's private key information is generated by calculating the commitment value and receiving a random challenge from the verifier. Based on the commitment value, the random challenge, and the response value, the correctness of the zero-knowledge proof equation is verified using the SM3 hash algorithm, and the authentication result is generated without exposing the user's private key.

8. The method according to claim 1, characterized in that, The deployment of the adaptive key update strategy involves acquiring network threat situation data, analyzing threat level changes through a threat situation assessment function, and obtaining dynamically adjusted key update parameters, including: Based on the authentication results and the number of abnormal network logins and attack frequency, a quantified threat level score is calculated using a threat situation assessment function to generate a multi-dimensional threat feature vector. Based on the multidimensional threat feature vector and historical threat data, an adaptive grid sampling strategy is used to perform dense sampling in areas where the threat level score exceeds a preset threshold, and a change point detection algorithm is used to identify key node information of threat situation changes, thereby determining the triggering time for key updates. Based on the threat level score and the key node information, the threat development trend is predicted through a multi-level threat threshold system and a local function approximation method, and the key update frequency and strength parameters are dynamically adjusted.

9. The method according to claim 8, characterized in that, The establishment of a secure communication channel based on SM4 symmetric encryption includes: Based on the dynamically adjusted key update parameters and the IBE mechanism, a temporary symmetric encryption key is established by generating a random session key and securely distributing it to both communicating parties using IBE encryption technology. Based on the temporary symmetric encryption key and the key distribution data to be transmitted, the data is encrypted using the SM4 symmetric encryption algorithm to obtain the encrypted data. Based on the encrypted data and the SM4-GCM authentication encryption mode, the integrity of the data is verified by calculating the message authentication code and attaching it to the encrypted data, thereby preventing tampering attacks during transmission and ensuring the security of data transmission during key distribution and update.

10. The method according to claim 1, characterized in that, It also includes blockchain-based decentralized key escrow solutions, including: Based on the decentralized key distribution mechanism, a consortium blockchain network is constructed and the key generation and distribution process is recorded on the blockchain, generating a key management blockchain record containing timestamps and digital signatures to ensure the transparency and immutability of the key management process; For the blockchain key management records, a smart contract is designed to automate the key lifecycle management, including periodic updates, permission changes, and emergency revocation operations, and to generate smart contract execution logs and state change records; Based on the smart contract execution log, the state change record, and the multi-signature mechanism, multiple authorized nodes are required to participate in key key operations. The blockchain consensus mechanism is used to verify the legality of key operations, establish a distributed trust system, and enhance the security of the decentralized key distribution mechanism.

11. The method according to claim 1, characterized in that, It also includes quantum-safe post-quantum cryptographic enhancement schemes, including: Based on the batch-updated user key and the NTRU lattice base algorithm, a quantum-resistant hybrid key system is constructed by integrating lattice-based cryptographic algorithms and encoding-based cryptographic algorithms, generating quantum-safe user private keys and a quantum threat assessment model; Based on the availability of quantum channels, the requirements of traditional key management, and the quantum-safe user private key, a quantum-safe key transmission mechanism is established by designing a fusion protocol of quantum key distribution and traditional key management, which prioritizes the use of quantum key distribution when quantum channels are available. Based on the quantum threat assessment model, the quantum-secure key transmission mechanism, and the current cryptographic algorithm strength, the quantum security of the batch-updated user keys is improved by dynamically adjusting the key length and encryption parameters of the quantum cryptographic algorithm.

12. The method according to claim 1, characterized in that, It also includes intelligent threat detection enhancement mechanisms based on federated learning, including: Based on the threat situation analysis results of the dynamically adjusted key update parameters, distributed threat detection nodes are deployed and threat identification models are trained collaboratively using federated learning technology to generate threat level assessment results and threat classification labels. Based on the threat level assessment results and the threat classification labels, a multi-level threat response mechanism is designed to automatically adjust the key update policy, access control policy and audit policy according to the threat level, and generate threat response execution records and system security configuration adjustment logs. Based on the threat response execution records, the system security configuration adjustment logs, historical attack patterns, and the current system status, predictive security analysis is performed and an artificial intelligence decision engine is integrated to predict potential security risks and deploy protective measures in advance.

13. An identity authentication system based on Chinese national cryptographic algorithms, characterized in that, include: The hierarchical key management module is used to establish a hierarchical key management system based on the national cryptographic IBE framework, obtain elliptic curve parameters and base point G, use the SM2 algorithm to generate system master key pairs and establish domain key generation centers for each identity domain, thereby realizing a decentralized key distribution mechanism. The privacy protection module is used to construct a domain-aware differential privacy protection module based on the decentralized key distribution mechanism, obtain the security level information of the user identity domain, define a privacy budget allocation strategy according to the security level information, add calibrated Laplace noise to the user identity feature vector, and generate privacy-protected identity authentication data. The batch key update module is used to design a distributed batch matrix multiplication protocol based on the privacy-protected identity authentication data, obtain key update requests from multiple users, decompose them to multiple computing nodes for parallel processing through secret sharing technology, and generate batch updated user keys. The zero-knowledge verification module is used to implement a zero-knowledge proof verification mechanism based on the batch-updated user keys. It constructs a zero-knowledge proof protocol through a commitment scheme based on the SM3 hash algorithm, and completes identity verification and generates authentication results without exposing the user's private key. The adaptive policy control module is used to deploy an adaptive key update policy based on the authentication result, acquire network threat situation data, analyze the changes in threat level through a threat situation assessment function, and obtain dynamically adjusted key update parameters. The secure communication module is used to establish a secure communication channel based on SM4 symmetric encryption according to the dynamically adjusted key update parameters, and to use a temporary session key for encryption to ensure the security of data transmission during key distribution and update.

Images