11 results about "Provable security" patented technology

The present invention discloses a system and method for secure information processing that guarantees data privacy, security, and auditability throughout the AI lifecycle. The system empowers the user to classify input data as "public" or "private," which dictates its routing into one of two segregated processing pathways. The public pathway handles non-sensitive information using external AI models, while the private pathway processes sensitive data in a secure, isolated environment. A novel federated embedding capability safely synthesizes insights from both pathways, enhancing the overall quality of the output through their combination, without exposing raw private data. Underpinning the architecture is a pervasive blockchain layer that provides an immutable, cryptographically verifiable record of every action, enforcing data handling policies and ensuring provable security and compliance for regulated industries.

The invention discloses a trusted digital space-oriented dynamic data sharing security method and system, and relates to the technical field of dynamic data sharing. The isomorphic design enables dynamic attributes such as aging scores and the like to be directly embedded into an access strategy to form a mathematical basis of a dynamic strategy, and when the data aging state changes, a promise value updated by a dynamic credible data label design module automatically triggers strategy re-evaluation of a CP-ABE data sharing module to realize cross-module penetration of the dynamic attributes; in a medical data sharing scene, data timeliness is confirmed to reach the standard and to be compliant with the identity of a visitor through single interaction, so that redundant overhead caused by multiple times of verification in a traditional scheme is avoided; according to the invention, the cross cooperation realizes the attribute-strategy-verification integrated control in the dynamic data circulation process, and when the data aging state changes, the smart contract automatically coordinates the label update and strategy re-evaluation, constructs a dynamic response closed loop on the cryptographic level, and provides a provable and safe mathematical base for the high-sensitivity data sharing.

The invention discloses a lightweight secure communication method, device and equipment for a power system and a medium, and provides a lightweight communication measure combining combined matrix certificateless key generation driven by a public key and layered seed caching based on time sequence-load joint scheduling. The key of the measure lies in that a short-term random material of a terminal, such as an accompanying public key / epoch, is coupled with a combined matrix index mechanism, so that provable forward security private key rotation is realized; a session encryption block is converted from static preset to time sequence-load combined adaptive scheduling and hierarchical cache, and a collaborative leakage response and cache security recovery mechanism is designed, so that provable security and efficient communication are considered under strict resource constraints. Therefore, the security, the high efficiency, the self-adaptability and the robustness of data transmission carried out by the power system resource-limited terminal are ensured.

The invention discloses an activation function security calculation method based on secret sharing and masks, which comprises the following steps: a first server and a second server flatten acquired fragment values into one-dimensional fragment vectors respectively, and mask calculation is carried out on the two fragment vectors respectively by taking random numbers as masks, the calculation result is sent to the third server; the third server performs mathematical reconstruction calculation according to a fragmentation calculation result, calculates an activation function value of a reconstruction result by using an activation function, performs arithmetic fragmentation on the activation function value by using a random number, and sends two arithmetic fragmentation to the two servers; and the two servers demask the received arithmetic fragments to restore the redistributed fragment values. According to the method, privacy data leakage can be prevented, and shared data certifiable security under the information theory meaning is achieved.

The invention discloses a heterogeneous high-performance distributed key agreement algorithm for supporting and resisting malicious participants in a block chain. The method is characterized in that a group negotiation intermediate message is generated under a heterogeneous cryptosystem between a certificate-based public key cryptosystem and a certificateless cryptosystem, and a cryptosystem conversion process is not needed, so that the performance overhead is reduced. All intermediate messages may be signed to verify integrity and used to identify, audit and defend malicious disruption and collusion attacks by participants. After key negotiation is completed, any encryptor can dynamically establish a session, customize a receiver group and generate a session key, and only a specified receiver can decrypt. According to the scheme, specific cryptographic operation and conversion rules are designed according to unique system composition and interaction processes in the heterogeneous system, and provable security of the heterogeneous cryptographic system is achieved.

The application discloses a lattice public key data encryption and decryption method and a key encapsulation method based on vector decoding. The method comprises the following steps: obtaining first target ciphertext data according to predetermined plaintext data, a first predetermined encryption method, a first target public key and a predetermined interference term, so that lattice public key data encryption can be realized; obtaining target plaintext data according to the first target ciphertext data, a first target private key and a predetermined decryption method, so that lattice public key data decryption can be realized. Meanwhile, the application also proposes a variant problem of a ring learning with errors problem (RLWE), namely, a subset-sum parity RLWE (sspRLWE) problem, and further optimizes the lattice public key data encryption method based on vector decoding in the application based on the variant problem. The public key encryption method and the key encapsulation method designed in the application have the characteristics and advantages of provable security, resistance to quantum computer attacks, short public key and ciphertext length, high calculation efficiency, low decryption failure rate, flexible parameter selection and the like.

The present application belongs to the technical field of post-quantum cryptography, and provides a provably secure post-quantum key encapsulation method based on LWR masking noise, a computer device and a storage medium. The method includes three stages, a key generation stage: generating a pair of asymmetric keys for the receiver of communication; an encryption stage: the sender encapsulates the shared secret using the public key of the receiver to form ciphertext; a decryption stage: the receiver uses the private key kept by itself to offset the public part in the ciphertext, and filters out the introduced rounding noise and masking noise, and restores the plaintext message encapsulated by the sender. The present application can realize compact provable security reduction from Module-LWR to Module-LWE without significantly increasing the computing overhead and bandwidth, and at the same time, eliminate the potential side channel leakage caused by the deterministic rounding structure.

The invention provides a lattice-based semi-quantum key distribution method and equipment. The method comprises the steps that a public key and trap door information are generated based on a generation algorithm of a lattice-based noisy trap door claw-free function, and a classic user side sends the public key to a quantum server; receiving image information, bit string information and measurement base information sent by the quantum server; performing inversion processing on image information based on an inversion algorithm of a lattice-based noisy trap door claw-free function and trap door information, calculating an exact state of a quantum state in a quantum server according to an inversion result, the image information and bit string information, and determining a session key of a classic user side based on the exact state of the quantum state and measurement base information; and the classic user side judges the correctness of the measurement base information and sends a correctness judgment result of the measurement base information to the quantum server, and the quantum server further derives a session key consistent with the classic user side. According to the embodiment, a quantum channel is not needed, and post-quantum provable security is achieved.

This invention discloses an AI-based autonomous learning and intelligent debugging method and system for reflow soldering processes, belonging to the field of manufacturing process control technology. The method is applied to a three-layer collaborative computing system consisting of an edge layer, a local layer, and a cloud layer. It includes: fusing design data of the target product with real-time process data collected by the edge layer; using the unified feature field as input, performing joint design and process optimization to generate a process control scheme verified by software security; distributing the process control scheme to the edge layer for execution, with a hardware arbitration module independent of the main control system providing physical-level security redundancy protection for the execution process; and uploading the updated knowledge to the cloud layer after production is completed. By constructing a fully autonomous closed loop, it achieves zero-sample rapid cold start for new products, joint global multi-objective optimization of design and process, dual provable security protection (software and hardware), and continuous autonomous evolution of swarm intelligence.

The application discloses a kind of power system lightweight security communication method, device, equipment and medium, the power system lightweight security communication method proposes a kind of combination matrix certificateless key generation driven by accompanying public key and the combination of layered seed cache based on timing-load joint scheduling lightweight communication measures, the key of this measure is to couple terminal short-term random material, such as accompanying public key / epoch and combination matrix index mechanism, to realize the forward security private key rotation of provable;And session encryption block generation is changed from static presetting to timing-load joint adaptive scheduling and layered cache, while designing collaborative leakage response and cache security recovery mechanism, so that provable security and efficient communication are considered under strict resource constraints, to ensure the security, efficiency, adaptability and robustness of data transmission carried out by power system resource-constrained terminal.

The application discloses an asymmetric resource text steganography method and system based on a double-scale language model, and belongs to the technical field of information security. The method comprises the following steps: generating a steganography text subsequence token by token based on a double-scale large language model; and extracting secret information token by token based on a first language model. Under the premise that the steganography security meets the provable security requirement, the application can realize adaptive adjustment of embedding capacity, reduce the calculation complexity of the decoding end, and thus improve the practicability and scalability of the steganography communication system.