26 February 2024 in IEEE Transactions on Intelligent Vehicles
Due to the dynamic and uncertain nature of the autonomous driving environment, traditional detection methods are ineffective in detecting complex and diverse anomalies. Additionally, existing detection methods relying on deep neural network may lack interpretability, making it challenging to infer the accuracy of the algorithms or the reasonableness of anomalies. To address these issues, this paper proposes an unsupervised anomaly detection method based on the physical invariant subspace. This method utilizes a Kalman Variational Autoencoder (KVAE) to learn the normal trajectory of vehicles and incorporates a subspace extraction (SSE) layer in the encoder to identify and capture discriminative subspaces relevant to the anomaly detection task in the latent space representation. The entire anomaly detection framework consists of a data preprocessing module, a KVAE with a SSE layer (SKVAE), an iteratively reweighted least squares state estimator, and an adaptive non-parametric cumulative sum anomaly detector. Firstly, the SKVAE separates spatial representation from action recognition and employs a Gaussian state space model to capture latent dynamic information in the sequential data, filtering out outliers and noise to enhance the accuracy and robustness of anomaly detection. Moreover, the adaptive non-parametric cumulative sum detector automatically adjusts the detection threshold to adapt to different distributions and variations in the anomaly data. This method demonstrates perfect detection accuracy and efficiency in numerical experiments, and the proposed detection model offers interpretability of the predicted results through the physical invariant characteristics and the extracted interpretable subspaces.
4 July 2025 in Computers & Security
6 December 2023 in Computers & Security
3 April 2025 in Future Generation Computer Systems
8 July 2025 in IEEE Transactions on Consumer Electronics
With the rapid advancement of 6G networks and artificial intelligence technologies, Space-Air-Ground Integrated Networks (SAGIN) are revolutionizing communication services. SAGIN seamlessly integrates satellite, Unmanned Aerial Vehicle (UAV), and terrestrial infrastructures to offer expansive coverage, ultra-low latency, and highly reliable connectivity. This paper introduces SACVTH, a novel chain-verifiable and trustworthy heterogeneous federated learning framework aimed at tackling critical challenges within the SAGIN context, including resource optimization, privacy protection, and security verification. In particular, we initially devise a sophisticated community partitioning mechanism tailored to task requirements and device capabilities to optimize resource utilization and ensure timely task execution. Subsequently, we propose an innovative lightweight verifiable secure aggregation protocol, that leverages sparse matrices and pseudorandom generators to compress verification data. Additionally, we implement a chain-based group broadcast mechanism that significantly reduces the frequency of direct communications and the volume of data transmissions between trainers and servers by facilitating intra-group model sharing and aggregation. This is achieved by enabling intra-group model sharing and aggregation, allowing trainers to communicate through intermediary relay nodes, thereby enhancing both communication efficiency and scalability within the federated learning framework. Rigorous security analysis demonstrates that SACVTH ensures the verifiability of aggregation results. Comparative evaluations with other state-of-the-art methods reveal that SACVTH protocol significantly outperforms others in resource optimization and security verification, leading to substantial improvements in the efficiency, robustness, and reliability of federated learning frameworks.
9 September 2025 in Computers & Security