ietf-asrg
[Top] [All Lists]

[Asrg] DCC and IP checksums

2003-03-10 09:12:05
From: Hadmut Danisch <hadmut(_at_)danisch(_dot_)de>

In other words: How much can messages differ to still get the same
hash and what's the minimum difference that guarantees a different
hash?

It doesn't matter.

ANY checksum scheme can be defeated, and usually very easily (unless
the checksum is useless for practical message distinction.)  It's
trivial to create N messages with arbitrarily-large Hamming (or
"spamming") distance.  Even without knowing the details of the
checksum algorithm, a few simple black-box experiments will probably
reveal its weakness.  If you know the algorithm, it's even easier.

That doesn't mean checksums are useless; on the contrary.  It is a
worthwhile cause to block spam from unsophisticated spammers.  Just
because we know we'll never completely stop determined spammers, it
doesn't mean we should give up on the low-hanging fruit.

The best we can hope for is to increase the hassle of spamming.  Every
extra hoop a spammer has to jump through means a certain percentage of
spammers will simply give up (or else fail because of lack of
hoop-jumping.)

--
David.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg