[Cryptography] Rubber-hose resistance?

[Jerry Leichter]

> What attack will filling the disk with random bits stop that simply zeroing the disk won't stop?  If you're worried about national-lab level attacks ... who really knows what they are? 
>  
> Zeros allow digging deeper into noise and are a 'simple' known.
> They present a signal that can be subtracted with more precision from the analog data stream than some.
> 
> The a pass of random followed by zeros would be difficult to dig out.
And you know this ... how?  The actual magnetization pattern on a modern disk has no simple relationship with the the bit pattern as seen all the way outside the disk.

Maybe all zeroes are easier to remove.  Maybe a random bit pattern is easier to remove.  Take a simple analogy:  If you were talking about audio, a random bit pattern would be like high-frequency hiss, easily removed by a low-pass filter.

The classic "35 patterns" paper carefully examined a number of long, long obsolete disk encodings and proposed specific patterns suitable for maximally erasing bits encoded with each particular technology.  There were 35 patterns because *different* erasure patterns, and different *combinations* of erasure patterns were most suitable.  If random bits were good enough, there would have been no paper to write.

I don't know what pattern are best for modern disks - or, for that matter, how this varies from disk type to disk type.  And neither do you.  Without careful study, simply asserting that random bits "make it more difficult" just won't do.

Of course, if you're talking SSD's, even ignoring remapping ... the pattern written is likely irrelevant, since each page has to be erased before it's written.

                                                        -- Jerry


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.metzdowd.com/pipermail/cryptography/attachments/20171220/ddac9e75/attachment.html>