Does SSL or HIP solve all TCP/IP security problem?

Question

I recently read this paper "A Look Back at Security Problems in the TCP/IP Protocol Suite" (https://www.cs.columbia.edu/~smb/papers/acsac-ipext.pdf) explaining few attacks at the TCP/IP stack.

One attack is at TCP layer (SYN flood, or DoS). Another is at the routing layer, i.e. RIP and BGP protocol. I understand that SSL and HIP provide encryption + authentication at level above IP layer, hence it doesn't really solve problem at the routing layer.

My question is that whether there are real/practical routing-based attacks when SSL or HIP are used, or are they merely academic problems? I mean, if the flow is encrypted and authenticated, what gain would the attacker have in being able to redirect or sniff the traffic?

What are the current techniques for securing routing protocols?

Answer 1

If "flow is encrypted and authenticated" the attacker is looking for a way to decrypt or break the authentication model.

Attacks that leverage MITM (man in the middle) methods use routing protocols (ARP and others) to put attackers 'between' the client and server where traffic can be captured and later unecrypted in some cases (Google WEP and Aircrack) or false authorization can be provided (Google SSLStrip where the client believes their connection is leveraging SSL/TLS, but in fact the connection is unencypted between the client and the attacker, then encrypted between the attacker and the 'secure' website.)

PS - Downgrade attacks can force web servers to use older and unreliable versions of SSL (like SSL3)... so unfortunately its not quite as easy as saying:

"what gain would the attacker have in being able to redirect or sniff the traffic?"