Pseudorandom number generators are deterministic algorithms capable of producing sequences of numbers that appear sufficiently ”random,” and they find extensive applications across various domains such as cryptography, network security, communications, machine learning, and gaming. As highly nonlinear mathematical systems, neural networks exhibit characteristics such as fitting ability, unidirectional property, generalization capability, and parallelism, which render them prominent in the design of PRNGs and a topic of significant research interest. To date, there exists no comprehensive review focusing on the utilization of neural networks for the design of PRNGs. This paper categorizes existing neural network-based PRNG design schemes into three types: those based on recurrent neural network models and their variants, such as Long Short-Term Memory (LSTM) models; those based on generative adversarial networks (GANs); and those based on deep reinforcement learning. Subsequently, the paper elucidates the design philosophies and typical algorithmic principles underlying these schemes, compares these algorithms, summarizes the existing challenges, and discusses prospective research directions.

Pseudo-random number generators (PRNGs) are deterministic algorithms that generate sequences of numbers approximating the properties of random numbers, which are widely utilized in various fields. In this paper, we present a Genetic Algorithm Optimized Generative Adversarial Network (hereinafter referred to as GAGAN), which is designed for the effective training of discrete generative adversarial networks. In situations where non-differentiable activation functions, such as the modulo operation, are employed and traditional gradient-based backpropagation methods are inapplicable, genetic algorithms are utilized to optimize the parameters of the generator network. Based on this framework, we propose a novel recursive PRNG. Given that a PRNG can be constructed from one-way functions and their associated hardcore predicates, our proposed generator consists of two neural networks that simulate these functions and serve as the state transition function and the output function, respectively. The proposed PRNG has been rigorously tested using stringent benchmarks such as the NIST Statistical Test Suite (SP800-22) and the Chinese standard for random number generation (GM/T 0005-2021). Additionally, it has demonstrated outstanding performance in terms of Hamming distance. The results indicate that the proposed GAN-based PRNG has achieved a high degree of randomness and is highly sensitive to variations in the input.

With the rapid advancement of technology, smart grid are increasingly replacing traditional grids. Smart grid offer efficient data management and analysis tools that significantly benefit users, but they also introduce new privacy challenges. Privacy-preserving data aggregation (PPDA) schemes have emerged as a solution, enabling the aggregation and processing of large-scale datasets without revealing any individual data points. This not only protects user privacy but also ensures data availability. In recent years, the field of smart grid privacy preservation has seen significant advancements in PPDA schemes. However, there has been a notable lack of comprehensive reviews that systematically summarize and analyze these developments. Our work aims to address this gap by providing a detailed survey of the latest PPDA schemes for smart grids over the past few years. This effort is intended to offer scholars a clear overview of the current state of the field and facilitate their understanding of the latest progress in smart grid privacy preservation. This paper provides a comprehensive survey of PPDA schemes for smart grid, analyzing their security requirements and discussing key implementation techniques. Based on whether the scheme relies on a trusted authority, we categorize the schemes into two main types. Schemes with a trusted authority are further divided into those that enhance increase data verifiable security, fault tolerance mechanisms, detect erroneous data, and support multi-type operations. In contrast, schemes without a trusted authority are classified into centralized architectures based on multiple nodes and decentralized architectures based on blockchain. This paper compares and analyzes the security, application scenarios, and performance of these schemes, and proposes future research directions based on the limitations of existing PPDA schemes.

Smart grid can be effectively enhanced by user data, however, data breaches pose significant risks related to privacy violations and data misuse. To address this issue, fault-tolerant privacy-preserving data aggregation schemes can protect data privacy while accommodating user drops. Most existing solutions depend on auxiliary computation in trusted authority for data aggregation. However, for practical smart grid applications, it is often unrealistic to expect these centers to remain online at all times. Designing a solution that does not rely on trusted authority (TA) while ensuring fault tolerance is a challenging task. In this paper, we propose an enabling efficient and fault-tolerant data aggregation scheme without TA, named EEFDA. Unlike existing encryption schemes, EEFDA supports reusable masks. We introduce the Reusable Shamir Threshold Secret Sharing (RSSS) technique, which allows users to securely adjust the secret without disclosing the initial sharing. The EEFDA scheme not only enhances system robustness but also reduces the communication overhead for inter-user interactions. Numerous experiments demonstrate that EEFDA communication is low-cost, efficient, and well-suited for resource-constrained devices.

Internet of Things (IoT) facilitates intelligent interconnection and data exchange between devices. However, existing data aggregation schemes face challenges such as edge node disconnections, high computational overhead, and limited fault tolerance, which affect the reliability and efficiency of the system. To address these issues, this paper proposes a smart contract assisted and fault-tolerant data aggregation scheme without a trusted authority, named Cap. To realize it, we integrate blockchain technology and homomorphic encryption. Through smart contracts, we dynamically select and switch edge nodes to ensure that the data aggregation process continues even if some nodes go offline. Moreover, by leveraging homomorphic secret sharing, we effectively minimize communication overhead and ensure that users can exit the aggregation process without jeopardizing data integrity. This scheme provides an efficient and reliable solution for secure data aggregation in Iot environments and enhances the robustness and data security of the system. Performance evaluation show that, compared with existing schemes, Cap reduces computational and communication overhead by 30%.

With the development of cloud computing, secure file storage based on it is even more widely popular. Attribute-based encryption (ABE) is one of the key cryptographic techniques to solve the security of cloud storage. China has independently developed its own cryptography industry standard - SM9 identification encryption, which will be wildly used in many areas. However, the current SM9-based ABE scheme cannot effectively utilize the parallelized resources of cloud computing. With the parallel computing technology, a parallel optimized ABE scheme based on SM9 is proposed. The scheme's accuracy and security are carefully examined, and its effectiveness is examined through tests.