Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

7 results about "Chaotic encryption" patented technology

A method for secure encryption of medical data

PendingCN122087844ADigital data protectionComputer hardwareKey space
This invention discloses a method for secure encryption of medical data, relating to the field of data encryption technology. The method includes implementing access control based on security levels, acquiring original medical images after authorization verification, stacking the images into a three-dimensional cube by channels, dynamically dividing them into blocks according to security levels, generating a chaotic encryption key based on the block information, and performing encryption operations on each three-dimensional block to obtain encrypted ciphertext. By destroying pixel correlation through channel stacking and segmentation, combined with differentiated keys generated by chaotic mapping, the key space and encryption efficiency are significantly improved, effectively resisting brute-force and statistical attacks. Permutation and substitution operations enhance the randomness of the ciphertext, ensuring encryption adapts to security levels and operates in a closed loop, guaranteeing controllable access permissions, improving the targeting and security of medical image encryption, and enabling flexible access for authorized users while ensuring efficient data encryption.
Owner:WUHAN AIYANBANG TECH CO LTD

An ai-based computer data risk identification system

PendingCN122346876ARisk levelCiphertext
This invention discloses an AI-based computer data risk identification system, belonging to the technical field of data analysis. It includes an encryption module, a training module, and an output module. The encryption module scrambles preprocessed image data using the Arnold scrambling operator and then encrypts the scrambled image data using a chaotic encryption method to obtain first data. The training module trains an attack triggering model based on the first data. The output module outputs the risk level based on the attack triggering model and the currently transmitted data. This invention constructs a collaborative encryption and analysis mechanism using the Arnold scrambling operator and a chaotic encryption method. It uses chaotic mapping to generate pseudo-random sequences to achieve numerical obfuscation, directly training the attack triggering model in the ciphertext domain. This allows for the identification of abnormal patterns without decryption, improving the efficiency and accuracy of risk identification.
Owner:NANTONG INST OF TECH

A probabilistic shaping encryption method, decryption method, encryption system, and decryption system based on symbol-level chaotic labels.

PendingCN122316602AAlgorithmCyclic prefix
This invention discloses a probabilistic shaping encryption method, decryption method, encryption system, and decryption system based on symbol-level chaotic tags in the field of optical communication transmission technology. The method includes: performing serial-to-parallel conversion on the signal to be transmitted to obtain a parallel signal; generating a first chaotic encryption sequence, a second chaotic encryption sequence, and a third chaotic encryption sequence using a three-dimensional chaotic system; performing symbol-level chaotic tag conversion on the parallel signal using the first chaotic encryption sequence to obtain a symbol-level encrypted signal; performing asymmetric polygon constellation mapping on the symbol-level encrypted signal to obtain an 8QAM constellation; performing symbol masking and subcarrier masking on the second and third chaotic encryption sequences to obtain a first scrambling matrix and a second scrambling matrix; scrambling the 8QAM constellation using the first and second scrambling matrices, performing an inverse Fourier transform, and adding a cyclic prefix to obtain an encrypted transmittable signal.
Owner:NANJING UNIV OF INFORMATION SCI & TECH

An adaptive multi-level digital watermarking method based on encoding process optimization

ActiveCN121940550BDecision modelAlgorithm
This invention discloses an adaptive multi-layer digital watermarking method based on encoding process optimization. The method first performs deep preprocessing on the input video stream using a pre-trained multimodal deep learning model to extract spatial complexity, temporal dynamics, and content saliency features, generating a unified feature vector. Based on the feature vector, a dynamic decision model is constructed to adaptively determine the watermark strength, embedding position, and type, achieving precise matching between watermark parameters and video content characteristics. A multi-layer watermark hierarchical embedding mechanism is employed, embedding robust invisible watermarks, fragile invisible watermarks, and dynamically visible watermarks into the video respectively, and combining chaotic encryption or hash chain encryption to enhance security. During copyright verification, an adaptive extraction algorithm recovers the watermark information of each layer, and a blockchain evidence storage system is used to complete data verification and infringement tracing. This invention constructs a complete copyright protection system encompassing embedding, encoding, extraction, and verification.
Owner:HANGZHOU BAOMIHUA TECH CO LTD

An artificial intelligence-based network security encryption authentication method and system

PendingCN122339742AAttackEngineering
This invention belongs to the technical field of network security encryption and authentication methods, and particularly relates to a network security encryption and authentication method and system based on artificial intelligence. It constructs an input layer security barrier to block adversarial sample attacks by employing dual-dimensional perturbation detection, cryptographic hash comparison, and feature desensitization based on terminal hardware identifiers. It also establishes a training data tracing and poisoning sample removal mechanism, coupled with backdoor feature detection at the inference end and incremental model updates, to resist data poisoning and backdoor implantation threats. Furthermore, it utilizes chaotic encrypted sharding storage, dynamic parameter updates, and query behavior fingerprint detection to prevent model theft and reverse engineering attacks by returning false results and shielding intermediate layer outputs. Finally, it establishes a closed-loop linkage between adversarial perturbation, data poisoning, and model theft protection modules, simultaneously strengthening defense strategies upon alarms, and achieving unsupervised adaptive optimization through data collection to adapt to new and unknown attacks.
Owner:HUNAN COLLEGE OF INFORMATION