tempest monitoring of key bit flips

proff@iq.org
Thu, 18 Jun 1998 23:35:12 +1000 (EST)

• Next message: Lakey, Jeremy: "RE: $100,000 reward"
• Previous message: Zachary Roger Amsden: "$100,000 reward"
• Next in thread: Simon R Knight: "Re: tempest monitoring of key bit flips"
• Reply: Simon R Knight: "Re: tempest monitoring of key bit flips"

>
> > This is actually a wider issue than data just being pages out, Peter
> > Gutmann wrote a paper on data erasure/data persistence which makes
> > for good reading. Check
> > http://www.cs.auckland.ac.nz/~pgut001/pubs/secure_del.html for more
> > details.
>
> Yes I have this paper ... it does make very good reading ! It was
> this paper, which introduced me to the possibility of bit flipping
> sensitive data in physical memory, but until data can be reliably
> lock/fixed/pinned in position, the technique can't be applied. Not
> that any government agency is going to crash in ... rip out my users
> motherboards, and quickly scan for fading bit remanence anyway.

It does make good reading. However, I have a problem with the idea
of continual bit-flipping key bits. It seems like a tempest monitoring
dream. I'd imagine that even for an incredibly very week signal,
if you monitored the rf given off for the memory (or cache) bus
and performed a simple statistical analysis on the results, you
could quite easily infer the key from the regular pattern of key
- key inverse.

Cheers,
Julian.

• Next message: Lakey, Jeremy: "RE: $100,000 reward"
• Previous message: Zachary Roger Amsden: "$100,000 reward"
• Next in thread: Simon R Knight: "Re: tempest monitoring of key bit flips"
• Reply: Simon R Knight: "Re: tempest monitoring of key bit flips"