Re: [v6ops] "The Internet is for End Users" (Re: I-D Action: draft-ietf-v6ops-unique-ipv6-prefix-per-host-07.txt)

[Tom Herbert <[e-mail]>]

On Thu, Aug 17, 2017 at 10:38 AM, Ted Lemon <[e-mail]> wrote:
> El 17 ag 2017, a les 13:19, Tom Herbert <[e-mail]> va escriure:
>
> If you want security and
> privacy it should be implemented end to end.
>
>
> Even that's not good enough, since your metadata can be tracked.
>
Ted,

That's a good argument that everything in a packet should be encrypted
except for the IP header and options that are necessary for delivery.
This becomes essential to prevent tracking. For instance, one of the
problems posed in IDEAS list is that long lived connections are easily
tracked in time. A proposed solution is to allow dynamically changing
addresses of an existing connection. This obfuscates the addresses
nicely, but if there's still connection acks #s, seq #s, and ports in
plaintext it's little bother to be able track the connection across an
address change. So to prevent such tracking, anything in the packet
that could be used for tracking identity needs to hidden or
obfuscated. Simply using random addresses is not nearly enough.

Tom