Search

Kernel methods are popular machine learning methods that provide automated pattern analysis of raw datasets. Of particular interest is Support Vector Machines that are used to solve supervised machine learning problems in many areas such as business, finance and healthcare. Nowadays, complex computations and data analytic tasks can be outsourced to specialized third parties. However, data owners might be reluctant to share their data especially when it includes sensitive information. Therefore, a need for privacy-preserving machine learning applications cannot be overstated. We present FHSVM: a Fast Homomorphic evaluation of non-linear SVM prediction on encrypted data using Fully Homomorphic Encryption. We provide design, implementation and several algorithmic and architectural optimizations such as novel packing strategies and parallel implementation to achieve real-time private prediction. We employed the CKKS FHE scheme to implement FHSVM under 128-bit security level. We evaluated FHSVM on a contemporary real-world large dataset compiled for anti-money laundering tasks in Bitcoin transactions. Empirical analysis demonstrates that homomorphic SVM prediction can be performed in 1.25 s on multi-core CPU platforms. In addition, FHSVM shows zero accuracy loss when compared to the non-privacy-preserving implementation. This shows that FHSVM is both computationally secure and fully utilizes the data. [ABSTRACT FROM AUTHOR]

Recently, Unmanned Aerial Vehicles (UAVs) have become a widely popular technology with remarkable growth and unprecedented attention. However, UAV communication networks are susceptible to various cyber-intrusions/threats due to their limited computation and communication capabilities. Such intrusions/misbehaviors tend to be processed as normal packets through the UAV communication networks. In this work, we present an autonomous intrusion detection system that can efficiently detect the malicious threats invading UAV using deep convolutional neural networks (UAV-IDS-ConvNet). Specifically, the proposed system considers encrypted Wi-Fi traffic data records of three types of commonly used UAVs: Parrot Bebop UAVs, DBPower UDI UAVs, and DJI Spark UAVs. To evaluate the developed system, we employed the UAV-IDS-2020 dataset which includes various attacks on UAV networks in unidirectional and bidirectional communication flow modes. Moreover, we emulate the context of homogeneous and heterogeneous networked UAVs. Our best experimental outcomes exhibited a victorious intrusion detection accuracy of 99.50% for the two-class classifier model (normal UAV vs. anomaly) with 2.77 ms prediction time. Besides, the proposed system was evaluated using other performance metrics including confusion matrix parameters, false alarm rate, detection precision, detection sensitivity, and prediction overhead. The performance analysis showed that our UAV-IDS-ConvNet system outperforms several recent existing intrusion detection systems developed to secure the UAV communication networks by (6–23) %. [ABSTRACT FROM AUTHOR]