The server has to do an encryption and a decryption. A legit client has to do the pre-SYN dance. In your protocol, I don’t spam pre-SYNs; I spam SYNs with fak

The server has to do an encryption and a decryption. A legit client has to do the pre-SYN dance. In your protocol, I don’t spam pre-SYNs; I spam SYNs with fake tokens. Thus I send only the same number of packets as before, but the server has to do these decryptions! Or I can mix and match spoofed SYNs with pre-SYNs and legitimate SYNs in whatever way finds the best leverage. Your pre-SYN design is very close to the design of TLS 1.2, by the way. That’s how I knew how to attack it. TLS is attackable in just this way: open a real connection, then lots of fake handshake requests. No crypto required to generate them, but the server has to do lots. Now that said, BCP 38 would go a LONG way towards addressing this. There will come a day when the well-behaved networks decide they’re not accepting traffic from non-compliant peers.