Aug 2026· 2026 International Conference on Intelligent Multimedia, Networking, and Security (IMNS)· pp. 1-6· 0 citations· 20 references
Abstract
Local differential privacy (LDP) can mask poisoning behavior in crowdsensing by making poisoned sensory data difficult to distinguish from benign sensory data, which in turn makes malicious workers difficult to identify and undermines reliable truth discovery. Conventional value based defenses lose discriminative power when privacy noise is large, whereas recent defenses designed for LDP settings rely on computationally expensive optimization or game theoretic modeling. To address this problem, we propose 4D-KDE, a lightweight defense framework. Specifically, 4D-KDE represents each worker using four residual statistics: mean bias, magnitude, skewness, and kurtosis. It then applies kernel density estimation to the resulting feature vectors, assigns anomaly scores to workers, and filters suspicious workers before aggregation. Experiments on a synthetic dataset and a real-world dataset show that 4D-KDE reduces attack gain and achieves up to 24× speedup over representative state-of-the-art baselines.
In many advanced recommender systems (e.g., GNN-based), local smoothing mechanisms would distill collaborative signals but could inadvertently amplify targeted data poisoning threats. Existing defenses predominantly relied on rigid binary filtration strategies. However, such structural deletion of suspicious nodes ofte...
Yunmeng Zhao, Jun Shen, Shenbao Yu et al.· Proceedings of the 32nd ACM...· 0 citations
Insider threats remain one of the most damaging and difficult-to-detect cybersecurity risks, as malicious actors already possess legitimate access privileges. This paper presents a privacy-preserving detection framework that combines per-user process mining with differential privacy, rigorously evaluated on the officia...
Walter Maanyere, Xian-Wen Fang, Palvine Ngob Enow et al.· International Journal of Sci...· 0 citations
Federated learning (FL) enables multiple clients to train a global model collaboratively without sharing their raw data. However, compromised clients can launch data poisoning attacks by submitting malicious local updates and thereby degrade model accuracy. In addition, unprotected model updates may disclose sensitive...
Jia-Yin Li, Shengmin Xu, Xing-Shuo Han et al.· IEEE Transactions on Network...· 0 citations
ProxyDrift is presented, a framework that identifies and measures drift between production traffic and offline evaluation sets, and constructs and refreshes those evaluation sets accordingly; all without access to raw user data.
Michael Levit, Josh Ledgard, Haoyu Dong et al.· 0 citations
A federated deep learning framework that systematically integrates adaptive privacy noise mechanisms and trust-weighted aggregation within a distributed architecture that ensures the protection of sensitive data during collaborative analysis through precise differential privacy control and advanced neural network model...
