FINDING · DETECTION

CLD-Net achieves 98% accuracy distinguishing VPN from non-VPN encrypted traffic on the ISCX public dataset, using an end-to-end CNN+LSTM model that requires no hand-crafted expert features.

From 2021-hu-cld-net-network-combining — CLD-Net: A Network Combining CNN and LSTM for Internet Encrypted Traffic Classification · Abstract · 2021 · Security and Communication Networks

Implications

VPN protocols are distinguishable from non-VPN traffic with near-certainty by commodity ML pipelines; circumvention tools should not rely on the encrypted-tunnel label alone to evade detection.
End-to-end classifiers that learn from raw payloads remove the feature-engineering bottleneck, meaning novel protocols are not automatically safe — any new transport should be evaluated against CNN-style spatial classifiers, not just signature-based tools.

CLD-Net achieves 98% accuracy distinguishing VPN from non-VPN encrypted traffic on the ISCX public dataset, using an end-to-end CNN+LSTM model that requires no hand-crafted expert features.

Implications

Tags