Mark Aldrich writes:
Teledyne (TET) is out marketing a "new" crypto system as a competitor against DES and what TET calls "linear" cryptosystems. They are claiming a non-linear approach in which (as best I can tell) they are permuting the s-boxes and using 4 bit/16 entry substitution tables.
They are also asserting that these "key generated substitution tables and inter-round permutations" are supported by "nonlinear orthomorphic mappings generated from arbitrary key" (and that this is patented) and "row-complete Latin Squares generated from an arbitrary key" (and that this is also patented).
Lothrop Mittenthal, who is Senior Cryptologist at TET, gave a sparsely- attended talk about some or all of this stuff at the RSA conference in January. The title was "Statistically Efficient Inter-round Mixing in Block Substitution Devices". I came in late and was ill-prepared to evaluate much of what he was presenting, so I can't offer any useful technical opinion. Copies of his transparencies are in the RSADSC proceedings, so you might try to borrow someone's copy, or I suppose someone could scan `em and mail you a copy. He didn't give an email address AFAIK, but according to the proceedings his phone is (in the PST zone) (805) 498-3621 ext. 5005 Lots of cypherpunks were wandering around there, so maybe someone else heard the whole seminar and can comment. (CME ?) -Lewis "You've got no secrets to conceal. How does it feel ?" (Dylan)