Abstract The advancement of quantum attacks has posed a significant threat to blockchain cryptosystems. Furthermore, the finite lifecycles of ciphertext and keys may result in data leakage risks for long-term storage due to potential breaches in ciphertext security, necessitating ciphertext updates. Ensuring the long-term security of cryptographic algorithms on blockchains against quantum attacks has emerged as a critical concern in blockchain research. A blockchain oracle serves as a mechanism for data communication. The Module-Learning with Errors (MLWE) problem is a significant issue in lattice-based cryptography, providing security against quantum attacks. To resist the threat of quantum attacks on encrypted data and ensure the long-term security, this study proposed an updatable encryption (UE) scheme for blockchain oracles based on post-quantum cryptography. First, this study developed a UE scheme based on the blockchain oracle architecture, enabling the oracle to update encrypted data via update tokens, ensuring the data's long-term security. Additionally, a UE algorithm based on the MLWE problem was introduced to enhance its resistance to quantum attacks. Finally, an error-correcting mechanism tailored for the UE algorithm was developed to validate transmitted data, correct errors, and improve the scheme's robustness. The experimental results demonstrate that the proposed scheme not only provides quantum-resistant security but also enables ciphertext updates through update tokens. Compared to existing schemes, it has more advantages in protecting data security.

Online user interactions produce privacy-sensitive data, necessitating data governance frameworks that respect user-defined privacy preferences. Nevertheless, the centralized storage and management of such preferences encounter significant interoperability challenges across different domains. By combining on-chain and off-chain storage paradigms, hybrid-storage blockchains have developed into a reliable infrastructure. Their value lies not only in ensuring secure data storage, but also in providing a practical foundation for data governance, especially in complex and cross-domain environments. In this paper, we propose LiveIndex, an index-driven hybrid-storage blockchain framework for cross-domain data governance. First, we design an on-chain data index based on distributed chameleon hash (DCH) to enable efficient off-chain data governance. This index incorporates essential elements, including metadata tags, data addresses, privacy ciphertexts, and group signatures. Second, we employ asymmetric searchable encryption (ASE) to protect privacy by converting analysis policies into search trapdoors and embedding privacy attributes into ciphertexts, thereby supporting compliant and secure data analysis. Third, we integrate proxy re-encryption (PRE) to strengthen control over cross-domain data, using re-encryption keys to ensure secure data authorization. Finally, we implement and evaluate LiveIndex on public datasets. The results show time overheads ranging from milliseconds to microseconds, demonstrating the practicality and efficiency in real-world scenarios.

To address security issues caused by limited computing resources, weak security protection, and sparse malicious sample data in IoT environments, this paper proposes a large language model-driven lightweight method for IoT malicious traffic detection (LLMD-IMTD). First, a large language model-driven small sample data enhancement algorithm is used to solve the problem of high detection false positive rate caused by too few malicious traffic samples, so as to alleviate the impact of uneven data distribution on malicious traffic detection performance. After that, a robust traffic feature selection algorithm based on meta-learning is used to select the feature subset that has the greatest impact on the malicious traffic detection task in a specific scenario, and the feature subset is dynamically updated according to the characteristics of the attack traffic, so as to quickly adapt to lightweight attack traffic detection tasks. Finally, combining the malicious traffic detection algorithm based on grid search and multi-classifier detection, the malicious traffic classifier is constructed, and the multi-model collaborative voting mechanism is used to predict the detection results, so that the algorithm can achieve more stable and accurate traffic detection under the complex network traffic environment. The experimental results show that the proposed method can achieve the F1 Score of 99.25 on the BoTIoT-2018 dataset, which has better detection performance than the benchmark method.

ABSTRACT Access control is a critical security measure to ensure that sensitive information and resources are accessed only by authorized users. However, attribute‐based access control in the big data environment faces challenges such as a large number of entity attributes, poor availability, and difficulty in manual labeling. In this paper, we focus on the problem of mining and optimizing security attributes of unstructured data resources and propose a method for mining security attributes of unstructured textual data based on multi‐model collaboration. First, we utilize unsupervised methods to extract candidate attributes from textual resources, and then weight the results of multiple methods using rough set theory to obtain the optimal result. Second, considering various factors including the text itself and the candidate attributes, we construct a feature vector consisting of 45 categories to represent the candidate attributes. Third, we employ a multi‐model voting method to collaboratively train the attribute mining model and obtain the security attributes of textual resources. Finally, based on HowNet, we optimize the security attributes to achieve automated and intelligent mining of access control data resource security attributes, providing an attribute foundation for precise access control. The experiments indicate that the attribute mining precision rate of the method proposed in this paper can reach up to 92.36%, F1‐score can reach up to 82.51%. The attribute scale can be compressed to 69.59% of its original size after optimization. This method has a greater advantage over other methods and can provide attribute support for access control of large data resources.

Abstract Microservice architecture faces security issues due to the large scale of services, resulting in a large attack surface. Integrating the concept of “continuous authentication, never trust” from the zero-trust security model into the access control of microservices has been extensively studied, but current microservice access control mechanisms still face issues like poor adaptability, inadequate security measures, and limited control capabilities, failing to fully implement zero-trust principles. In response, this paper proposes a dynamic trust evaluation and access control model named GATAC, based on machine learning and a game-theoretic reward-punishment algorithm. To meet the need for dynamic and highly accurate access evaluation, the model first introduces a trust computation method that enhances the FORBP neural network using random forest. Furthermore, to improve the access control capability of microservices, a game state based reward and punishment algorithm (MICGA) is proposed to encourage honest user access behaviors. Experimental results indicate that the proposed solution outperforms existing access control mechanisms in terms of accuracy, resistance to attacks, and adaptability.

Abstract Federated learning is a promising new distributed machine learning paradigm, where the client realizes secure and collaborative multi-user training of machine learning models by retaining private data and sharing model parameters with the server. However, with the frequent interaction of model parameters between the client and the server, the client will consume a large amount of network and arithmetic resources, and resource-constrained clients can hardly maintain model security while ensuring the efficiency of collaborative user training. Therefore, we propose FedRtid, a shuffle differential privacy federated learning scheme with random participation and adaptive time constraints, to improve the efficiency of collaborative user training while considering model privacy. First, in model training, the participating clients have the right to decide on random participation in training locally and independently, to alleviate the user’s resource constraints and reduce the time of user interaction to train the model, while adding differential noise to the shared model parameters to ensure model security. In addition, to avoid the global model security decline of server aggregation due to fewer clients participating in training, and the model accuracy decline caused by adding differential noise to all model parameters, we constructed user sparsification and adaptive time-constrained shuffle techniques to reduce the number of model parameters to which the user adds noise, and enhance the model security. Under two types of data distributions, independently and identically distributed and non-independently and identically distributed, we conduct a large number of experiments on three real datasets, and the results show that FedRtid can effectively balance the accuracy and privacy of the model.

Access control technology is critical to the safe and reliable operation of information systems. However, owing to the massive policy scale and number of access control entities in open distributed information systems, such as big data, the Internet of Things, and cloud computing, existing access control permission decision methods suffer from a performance bottleneck. Consequently, the large access control time overhead affects the normal operation of business services. To overcome the above-mentioned problem, this paper proposes an efficient permission decision engine scheme based on machine learning (EPDE-ML). The proposed scheme converts the attribute-based access control request into a permission decision vector, and the access control permission decision problem is transformed into a binary classification problem that allows or denies access. The random forest algorithm is used to construct a vector decision classifier in order to establish an efficient permission decision engine. Experimental results show that the proposed method can achieve a permission decision accuracy of around 92.6% on a test dataset, and its permission decision efficiency is significantly higher than that of the benchmark method. In addition, its performance improvement becomes more obvious as the scale of policy increases.

Safe and scalable dynamic autonomous data interaction between medical institutions can increase the number of clinical trial records, which is of great significance for improving the level of medical trial collaboration, especially for clinical decision-making with regard to rare diseases. Through a preset authorization access and consensus mechanism, consortium chain provides integrity and traceability management for medical clinical data. However, how to enable users have ownership of their own medical data and share their medical data safely and dynamically between different medical institutions remains an area of particular concern. To achieve dynamic communication between medical consortium chains, this paper proposes (i) a cross-chain communication mechanism by simplifying the heterogeneous node communication topology and (ii) the construction rules of the node identity credibility path-proof to carry out dynamic construction and verification of the path-proof for cross-chain transactions. In addition, the consensus of the cross-chain transaction is modeled as a threshold digital signature process with multiple privileged subgroups; thus, the intra-chain consortium consensus based on the verification node list is extended to the cross-chain consensus. A smart contract deployment and execution scheme based on rational node value transfer mechanism is proposed by analyzing the value transfer game between nodes. Experimental results showed that the proposed scheme can not only enable patients to share their records safely and autonomously in an authorized medical consortium chain within milliseconds but also realize dynamic adaptive interaction among heterogeneous consortium chains.