Search

Floating offshore wind platform (FOWP) has become the economically favored option for supporting wind turbines in deep waters. It is urgent to propose new concept designs for FOWPs that can be effectively deployed. Additionally, the extensive use of steel in such platforms significantly escalates costs, necessitating the optimization of steel utilization. Motivated by these challenges, a V-shaped floating semi-submersible platform equipped with NREL 5 MW wind turbine is designed and analyzed based on the potential flow theory and the blade element momentum theory. Fully coupled time-domain simulations are conducted using the F2A program, which couples NREL FAST and ANSYS AQWA via a Dynamic Link Library (DLL), to compare the hydrodynamic performance and stability of the V-shaped floating platform with the original triangle-shaped model of "Fuyao". Various sea conditions have been considered, including combined wind-wave action and wind-wave-current action at different incidence angles. The results show that the V-shaped floating platform has better economic and hydrodynamic performance (e.g., a reduction of 40.4% and 12.9%, respectively, in pitch and yaw motions, and a 17.4% reduction in maximum mooring tension), but lower stability than its triangle-shaped counterpart. [ABSTRACT FROM AUTHOR]

In this paper, we study the problem of privacy-preserving data synthesis (PPDS) for tabular data in a distributed multi-party environment. In a decentralized setting, for PPDS, federated generative models with differential privacy are used by the existing methods. Unfortunately, the existing models apply only to images or text data and not to tabular data. Unlike images, tabular data usually consist of mixed data types (discrete and continuous attributes) and real-world datasets with highly imbalanced data distributions. Existing methods hardly model such scenarios due to the multimodal distributions in the decentralized continuous columns and highly imbalanced categorical attributes of the clients. To solve these problems, we propose a federated generative model for decentralized tabular data synthesis (HT-Fed-GAN). There are three important parts of HT-Fed-GAN: the federated variational Bayesian Gaussian mixture model (Fed-VB-GMM), which is designed to solve the problem of multimodal distributions; federated conditional one-hot encoding with conditional sampling for global categorical attribute representation and rebalancing; and a privacy consumption-based federated conditional GAN for privacy-preserving decentralized data modeling. The experimental results on five real-world datasets show that HT-Fed-GAN obtains the best trade-off between the data utility and privacy level. For the data utility, the tables generated by HT-Fed-GAN are the most statistically similar to the original tables and the evaluation scores show that HT-Fed-GAN outperforms the state-of-the-art model in terms of machine learning tasks. [ABSTRACT FROM AUTHOR]

Cloud platforms allow administrators or management applications with privileged accounts to remotely perform privileged operations for specific tasks, such as deleting virtual hosts. When privileged accounts are leaked and conduct dangerous privileged operations, severe security problems will appear on cloud platforms. To solve these problems, researchers focus on auditing privileged users' behaviors. However, it is difficult to automatically audit fine-grained privileged behaviors for graphical operating systems. Moreover, it is hard to prevent users from bypassing the audit system or to prevent hackers from attacking audit system. In this paper, we propose a Secure and Automatic Behavior Audit system named SA-UBA. It provides advanced deep learning models to automatically achieve fine-grained user behavior audits for graphical operating systems. Furthermore, it adopts cryptography-based account storage and sharing methods to securely manage privileged accounts. In particular, privileged accounts cannot be leaked even if SA-UBA is compromised by attackers. We built a threat model of a cloud platform to evaluate the security of the SA-UBA and conduct extensive experiments with SA-UBA in real scenarios. The results show SA-UBA introduces a small overhead on securely managing privileged accounts and accurately recognizes fine-grained user behaviors. [ABSTRACT FROM AUTHOR]

The trust computing mechanism has an increasing role in the cooperative work of wireless sensor networks. However, the computing speed, resource overhead, and anti-collaborative attack ability of a trust mechanism itself are three key challenging issues for any open and resource-constrained wireless sensor networks. In this study, we propose a fast, resource-saving, and anti-collaborative attack trust computing scheme (FRAT) based on across-validation mechanism for clustered wireless sensor networks. First, according to the inherent relationship among three network entities (which are made up of three types of network nodes, namely base stations, cluster heads, and cluster members), we propose the cross-validation mechanism, which is effective and reliable against collaborative attacks caused by malicious nodes. Then, we adopt a fast and resource-saving trust computing scheme for cooperation between between cluster heads or cluster members. This scheme is suitable for wireless sensor networks because it facilitates resource-saving. Through theoretical analysis and experiments, the feasibility and effectiveness of the trust computing scheme proposed in this study are verified. [ABSTRACT FROM AUTHOR]

In this paper, we study the problem of privacy-preserving data synthesis (PPDS) for tabular data in a distributed multi-party environment. In a decentralized setting, for PPDS, federated generative models with differential privacy are used by the existing methods. Unfortunately, the existing models apply only to images or text data and not to tabular data. Unlike images, tabular data usually consist of mixed data types (discrete and continuous attributes) and real-world datasets with highly imbalanced data distributions. Existing methods hardly model such scenarios due to the multimodal distributions in the decentralized continuous columns and highly imbalanced categorical attributes of the clients. To solve these problems, we propose a federated generative model for decentralized tabular data synthesis (HT-Fed-GAN). There are three important parts of HT-Fed-GAN: the federated variational Bayesian Gaussian mixture model (Fed-VB-GMM), which is designed to solve the problem of multimodal distributions; federated conditional one-hot encoding with conditional sampling for global categorical attribute representation and rebalancing; and a privacy consumption-based federated conditional GAN for privacy-preserving decentralized data modeling. The experimental results on five real-world datasets show that HT-Fed-GAN obtains the best trade-off between the data utility and privacy level. For the data utility, the tables generated by HT-Fed-GAN are the most statistically similar to the original tables and the evaluation scores show that HT-Fed-GAN outperforms the state-of-the-art model in terms of machine learning tasks.