VPN obfuscation
Updated 21 September 2026 · How we research
Obfuscation hides the fact that you are using a VPN. Encryption protects the contents of your traffic; obfuscation attacks the shape of it, so that deep packet inspection cannot classify the connection in the first place.
How VPN traffic gets identified
- Port. The crudest signal. Blocking UDP 500, 4500 and 1194 stops IKEv2 and default OpenVPN outright.
- Handshake fingerprint. WireGuard, OpenVPN and IKE each have a recognisable opening exchange that matches regardless of port.
- Packet timing and size. Tunnels produce distinctive patterns even when contents are opaque.
- Active probing. The network connects to your server itself to see whether it behaves like a VPN endpoint.
Common techniques
All obfuscation trades speed for resistance to detection.
| Technique | Idea | Limitation |
|---|---|---|
| Run on port 443 | Use the port every network must allow | Defeats port blocking only; DPI still sees the handshake |
| TLS wrapping | Put the tunnel inside a genuine TLS session | Adds overhead and often forces TCP |
| Obfuscation layers | Scramble the handshake so it matches no known signature | Vulnerable to active probing if the server answers unusually |
| Proprietary transports | Vendor-specific designs such as ZoogVPN's Shadow and ZoogTLS | Shadow has no public specification; ZoogTLS is vendor-described as OpenConnect-based but its implementation is not published. Neither has been independently assessed. |
Nothing here is guaranteed
ZoogVPN's options
ZoogVPN publishes two in-house transports aimed at restricted networks: Shadow, listed for Windows, macOS, iOS and Android, and ZoogTLS, listed for Windows only. Both are described by the vendor as moderate speed, which is consistent with how obfuscation works. Shadow has no public specification; ZoogVPN describes ZoogTLS as based on OpenConnect with TLS encryption, which is at least a documented starting point.
Legal and personal risk