BGA Framework: Detecting Malicious Commands in Encrypted Traffic
量子位 · wechat · 2026-08-10
A research team from Beihang University proposed the BGA framework to tackle the challenge of detecting malicious attacks hidden in TLS 1.3 encrypted traffic. The high-entropy random noise of encrypted flows typically causes "attention dilution" in AI models, making it difficult to distinguish actual hacker signatures.
The BGA framework reconstructs the detection process across three dimensions: the data layer uses WGAN-GP to synthesize high-fidelity minority samples; the feature layer applies ANOVA to pinpoint core industrial control parameters; and the model layer introduces an Adaptive Gated Multi-Head Attention mechanism to suppress noise dynamically. Experiments show BGA achieves over 95.2% accuracy on complex IIoT encrypted traffic, with an inference latency of just 1.69ms on a single-core ARM environment, proving its viability for industrial edge gateway deployment.
More from Safety
- Kynx Stealer Malware Now Targeting AI and Dev Tool Credentials — TechNadu · 2026-08-10
- When Agents Supply Their Own Premises: Can Crypto Solve the Agent Trust Boundary? — docybo · 2026-08-10
- AI Replacing 1,000 Workers Exposes the Absurdity of Current Machine Tax Policies — VraserX · 2026-08-10
- NeurIPS Reviewer Complains: Most Papers Are Now AI-Generated Slop — HildeKuehne · 2026-08-10
- Raptor: Open-source project turns Claude Code into autonomous security agent — tom_doerr · 2026-08-10
- Google Overhauls Hacker Group Naming System with Country Codes — TechNadu · 2026-08-10