Fiber-optic distributed acoustic sensor (DAS) offers great potential for railway event monitoring due to its high sensitivity and robustness in complex environments. However, accurate recognition of acoustic events remains challenging, under real-time constraints where only short-duration signal segments with limited discriminative information are available. To overcome this, an efficient recognition framework was proposed by integrating multi-encoding image fusion and a lightweight transformer-convolutional network (CNN). Specifically, 0.032-second DAS signal segments were converted into complementary image representations, which were then fused into RGB images to enhance feature diversity. These multimodal images were processed by a compact backbone that combined vision transformer (ViT) modules with convolutional components, enabling effective extraction of multi-scale local and global features. The hybrid framework effectively mitigated the lack of temporal information in short segments, achieving 98.19% accuracy with only 1.33 million parameters and 3.3 ms latency per sample, substantially faster than existing methods. Compared to baseline models, the proposed architecture achieved superior accuracy and compactness, making it highly suitable for real-time and edge deployment in DAS-based monitoring systems.

Due to the loose authentication requirement between access points (APs) and clients, it is notoriously known that WLANs face long-standing threats such as rogue APs and network freeloading. Take the rogue AP problem as an example, unfortunately encryption alone does not provide authentication. APs need to be equipped with certificates that are trusted by clients ahead of time. This requires either the presence of PKI for APs or other forms of pre-established trust (e.g., distributing the certificates offline), none of which is widely used. Before any strong security solution is deployed, we still need a practical solution that can mitigate the problem. In this paper, we explore a non-cryptographic solution that is readily deployable today on end hosts (e.g., smartphones and laptops) without requiring any changes to the APs or the network infrastructure. The solution infers the Carrier Frequency Offsets (CFOs) of wireless devices from Channel State Information (CSI) as their hardware fingerprints without any special hardware requirement. CFO is attributed to the oscillator drift, which is a fundamental physical property that cannot be manipulated easily and remains fairly consistent over time but varies significantly across devices. The real experiments on 23 smartphones and 34 APs (with both identical and different brands) in different scenarios demonstrate that the detection rate could exceed 94%.

Motion sensors, especially accelerometers, on smartphones have been discovered to be a powerful side channel for spying on users' privacy. In this paper, we reveal a new accelerometer-based side-channel attack which is particularly serious: malware on smartphones can easily exploit the accelerometers to trace metro riders stealthily. We first address the challenge to automatically filter out metro-related data from a mass of miscellaneous accelerometer readings, and then propose a basic attack which leverages an ensemble interval classifier built from supervised learning to infer the riding trajectory of the user. As the supervised learning requires the attacker to collect labeled training data for each station interval, this attack confronts the scalability problem in big cities with a huge metro network. We thus further present an improved attack using semi-supervised learning, which only requires the attacker to collect labeled data for a very small number of distinctive station intervals. We conduct real experiments on a large self-built dataset, which contains more than 120 h of data collected from six metro lines of three major cities. The results show that the inferring accuracy could reach 89% and 94% if the user takes the metro for four and six stations, respectively. We finally discuss possible countermeasures against the proposed attack.

In the problem of privacy-preserving collaborative data publishing, a central data publisher is responsible for aggregating sensitive data from multiple parties and then anonymizing it before publishing for data mining. In such scenarios, the data users may have a strong demand to measure the utility of the published data, since most anonymization techniques have side effects on data utility. Nevertheless, this task is non-trivial, because the utility measuring usually requires the aggregated raw data, which is not revealed to the data users due to privacy concerns. Furthermore, the data publishers may even cheat in the raw data, since no one, including the individual providers, knows the full data set. In this paper, we first propose a privacy-preserving utility verification mechanism based upon cryptographic technique for DiffPart-a differentially private scheme designed for set-valued data. This proposal can measure the data utility based upon the encrypted frequencies of the aggregated raw data instead of the plain values, which thus prevents privacy breach. Moreover, it is enabled to privately check the correctness of the encrypted frequencies provided by the publisher, which helps detect dishonest publishers. We also extend this mechanism to DiffGen-another differentially private publishing scheme designed for relational data. Our theoretical and experimental evaluations demonstrate the security and efficiency of the proposed mechanism.