Outline for April 16, 1997
- Greetings and Felicitations
- Remember, talks about what you could do are at the end of this week
- History
- IBM did Lucifer, submitted it in response to NIST CFP
- NIST (really, NSA) suggested some minor changes; major one was to make key
56 bits, not 112.
- Show the cipher
- Product cipher with 64 bits in, 64 bits out, and 16 48-bit round keys
generated from 56 bit key
- Note S-boxes are real heart of algorithm
- Known attacks and weaknesses
- Complementation property: DESk(m) = (DESk'(m'))' where
x' is the bitwise complement of x;
- Weak, semiweak keys
- If it's a group, multiple encipherment worthless (as group is closed under
composition)
- differential cryptanalysis: first version unusable as at 16 rounds, more
plaintext/ciphertext pairs needed than exhaustive key trial; but for 15 rounds,
cuts this time. Later versions cut it to 247 tries. Works by comparing xors of
results with xors of corresponding plaintext.. Designers of DES knew about this
one, hence the design of the S-boxes
- linear cryptanalysis drops required chosen plaintext/ciphertext pairs to
242; not known to designers of DES.
- DES Modes
- ECB
- CBC
- note that OFB and CFB exist, essentially use DES as a pseudorandom bitstream
generator; OFB feeds back before xor, CFB after
- Triple DES and EDE mode
Notes by Elizabeth Jurrus:
[Postscript]
[Text]
[Microsoft Word]
You can get this document in
Postscript,
ASCII
text,
or
Framemaker
version 5.1.
Send email to
cs253@csif.cs.ucdavis.edu.
Department of Computer Science
University of California at Davis
Davis, CA 95616-8562
Page last modified on 5/12/97