1996-01-30 - Re: PPP link encryption?

Header Data

From: “Perry E. Metzger” <perry@piermont.com>
To: Alexandra Griffin <acg@mandrake.cen.ufl.edu>
Message Hash: 37ae293c0c5bbfa947fb9245352313e5aa2083329a1ea94d3999c8635fc7035f
Message ID: <199601301438.JAA02927@jekyll.piermont.com>
Reply To: <199601301136.GAA02305@mandrake.cen.ufl.edu>
UTC Datetime: 1996-01-30 15:54:29 UTC
Raw Date: Tue, 30 Jan 1996 23:54:29 +0800

Raw message

From: "Perry E. Metzger" <perry@piermont.com>
Date: Tue, 30 Jan 1996 23:54:29 +0800
To: Alexandra Griffin <acg@mandrake.cen.ufl.edu>
Subject: Re: PPP link encryption?
In-Reply-To: <199601301136.GAA02305@mandrake.cen.ufl.edu>
Message-ID: <199601301438.JAA02927@jekyll.piermont.com>
MIME-Version: 1.0
Content-Type: text/plain

Alexandra Griffin writes:
> Is there any software out there on the net for doing real-time,
> transparent encryption of a PPP link?

Well, there is software for encrypting IP datagrams. There may be PPP
encryption out there, since there is now an IETF protocol for doing
link encryption for PPP, but I don't follow PPP very closely.

> Also, if this link is running at 600kbit/s to 1Mbit/s, how much
> processing power would be required to keep up (assuming we're using,
> say, the IDEA cipher)?  Would a dedicated 386DX/33-based router on
> each end be sufficient?

I suspect not, but on the other hand hardware to do this sort of thing
is available. The TIS people packaged up swIPe with a 3DES board for
some applications a while back, just as an example of concept.

> Finally, could someone make a ballpark estimate as to the amount of
> additional latency that would be added?

Hard to say. Depends on the implementation, your link characteristics,