Slashdot Mirror
Inventors of RSA win Turing Award
Frisky070802 writes "The NY Times has an article on how Rivest, Shamir, and Adleman (the inventors of the eponymous RSA public-key encryption algorithm) have won the ACM's Turing award for contributions to computer science. You mean they didn't win already?"
I've always found it interesting that RSA is so hyped, and elliptic curve cryptography (ECC) -- introduced independently by Koblitz and Miller in 1986 -- has received so little attention.
... for an idea that's been around for more than 15 years, and which has made a lot of inroads in academia, it sure hasn't gone too far in industry (Certicom and Motorola being two notable exceptions).
Basically, with a 160 bit key, ECC achieves the same level of security as 1024 bit RSA. (Another example: a 591 bit ECC key is said to possess the same security as 15000+ bit RSA. Hyperelliptic curves may achieve comparable security in even smaller keys, but they remain mostly theoretical due to poor performance and questionable security relative to elliptic curves.) And, because the keys are so much smaller -- and there are efficient techniques for scalar point multiplication over a Galois Field -- ECC tends to offer incredibly good performance.
Unfortunately, the general adoption of ECC just isn't there in the same way as RSA. Mind you, there has been some recent research in which ECC was implemented in OpenSSL (0.9.6b), but I don't think that the SSL 3.0 protocol has been extended to support it yet. Also, there is a tremendous amount of research continuing in this field to improve ECC performance (not just in software, but in elliptic curve processors for smart cards, for instance). And, of course, ANSI, IEEE, and NIST FIPS 180-2 standards are working their way thru the pipes.
But still