The paper applies analytical modeling techniques to quantitatively analyze the reliability of the two-stage electric-vehicles (EV) charging service of State Grid e-charging platform (SGecP), one of the top 5 EV charging platforms in China. Reliability means that the SGecP EV-charging service satisfies performance requirements of EV-charging requests from users. We first develop a monolithic semi-Markov process (SMP) model for reflecting the dynamics of components involved in the two-stage service. To tackle the scalability weakness inherent in the monolithic model for a large-scale system, we then develop a novel hierarchical SMP model, which can effectively study the dynamics of SGecP charging service. We also derive the formulas for computing evaluation metrics. Simulation and numerical results validate the accuracy and efficiency of the proposed models, and highlight their potential to improve service reliability and user satisfaction.

In the cloud–edge–end collaborative architecture, when data traverses from Data Source Nodes (DSNs) to the Cloud Center Node (CCN) via Edge Verification Nodes (EVNs) for data aggregation, it is crucial to guarantee data integrity for the reliability of subsequent data analysis. However, traditional data integrity verification schemes suffer from high computation overhead, complex certificate management, and limited privacy protection. This paper aims to address these issues. We propose a Two-Tier Batch Data Integrity Verification (T2BDIV) scheme for the cloud–edge–end architecture. The scheme consists of three mechanisms. The first is the batch verification mechanism, which at the edge layer enables an EVN to efficiently verify integrity of data uploaded by DSNs before forwarding them to the CCN. Additionally, at the cloud layer, this mechanism enables the CCN to efficiently verify integrity of aggregated data from each EVN. The second is an identity-based signature mechanism to avoid the complexity of certificate management. The third is a dynamic Pseudo-Identities (PIDs) update mechanism, which generates dynamic pseudo-identities for each transmission to achieve anonymity and unlinkability. Security analysis demonstrates that the proposed scheme ensures data integrity, data source authenticity, and privacy preservation. Performance evaluation shows that our scheme significantly reduces computation and communication overhead compared to the existing scheme, making it highly suitable for large-scale cloud–edge–end deployments.

Attack-path planning plays a key role in proactive cybersecurity because of its ability in helping defenders anticipate adversaries and uncover critical vulnerabilities. This paper proposes GAPPO, a novel deep reinforcement learning-based attack path planning scheme that integrates Graph Attention Networks (GAT) and expert knowledge into Proximal Policy Optimization (PPO). There are three mechanisms in GAPPO. The first is using GAT to produce graph-structure-aware embeddings that emphasize critical connections, enabling expressive state representations for decision making. The second is a ruled-based action masking mechanism, which incorporates expert knowledge to prune the action space based on node dependencies and then to prevent illegal actions from negatively impacting training. The third is combining the results of the first two mechanisms into PPO for attack path planning. Our extensive experimental results demonstrate that GAPPO outperforms existing methods in terms of faster convergence and higher-quality attack paths across diverse scenarios.