This article can also be found in the Premium Editorial Download "Information Security magazine: Keep today's threats close and tomorrow's closer."
Download it now to read this article plus other related content.
|Gen 2 vs. Gen 3|
|By the time the Auto-ID
Center at MIT released EPC Generation 1 (and set up the EPCglobal standards body), many in the RFID
industry were already talking about Generation 2. Perhaps that's one reason co-founder Kevin Ashton
is confident that he will rally technologists and EPCglobal board members, including Sanjay Sarma,
to his proposed EPC Generation 3 standard.
EPC Gen 2 does more to improve on privacy than security, says Ashton. EPC Gen 2 includes a kill command, for example, with a 32-bit tag-specific password that most cryptographers agree can be picked up via a side channel attack.
EPC Gen 3 tags might include stored encrypted serial numbers and tag-and-reader authentication. Such measures would foil would-be tag counterfeiters and operators of rogue reader devices.
Once again, Ashton, a former brand manager for Proctor & Gamble, has his critics. But this time, the ACLU and end-time Christians are not among them. Rather, it is Ashton's RFID industry colleagues who are objecting to the EPC Gen 3 proposal, which he and his ThingMagic co-workers made in a recent EPC security whitepaper (http://thingmagic.com/ html/pdf/generation%202%20-%20security.pdf). (See "Gen 2 vs. Gen 3")
"P&G is wholly satisfied with Gen 2, and discussion about the evolution of Gen 3 at this time is misplaced and premature," P&G spokesman Paul Fox wrote in an email. Fox called the threats to most RFID deployments "theoretical."
Other retailers and their suppliers, at least for the moment, apparently consider the security provided by EPC Gen 2 tags to be adequate for their needs.
"So far, we did not experience any problems with hacks or comparable attacks," says Christian Maas, spokesman for European retailer Metro AG, also via email. "We are applying EPC Gen 2 standard in our logistical processes, which is secured in several ways, for instance, random number masking."
Random number masking is an EPC Gen 2 feature that adds a random number to a tag's ID to deter eavesdropping, and requires the tag and reader to exchange a digital handshake before they can exchange any data. The aim is to lock a tag so that only an authorized interrogator can write any data to it. But Ashton and others feel the random number masking is ineffective against a side-channel attack because the number is not encrypted.
This was first published in January 2007