At 01:32 AM 3/6/02 +0200, Sampo Syreeni wrote: >A while ago I thought about how one might apply public key signatures to >physical currency. It came to me that if there was a way to generate >measurable properties in the notes which couldn't easily be reproduced >(essentially, physical randomness in the notes' composure), it'd be >possible to make counterfeit notes detectable by signing a hash of a >measurement of such randomness by a well known public key and barcoding it >on the bill: if one possessed the key, one could produce arbitrary looking >bills and just sign the hash. If one didn't have the key, and couldn't >cheaply enough reproduce whatever gave rise to the measurement, one >couldn't reproduce a valid-looking note. The randomness could come from
>things like turbulent deposition of magnetic/fluorescent materials on the >bill, something one might assume to be fairly expensive to reproduce >exactly. Nice. >Is there some "prior art" to be found, anywhere? Does the idea sound >reasonable? (I tend to be of the opinion that paper money is here to stay >for quite a while, and improving on the state of the art of >anti-counterfeit measures seems like a fairly good idea. Over that >background, Chaumian cash and the like do not make a lot of a difference.) Yes, some company announced a way to "sign" chips via what IIRC was randomness in resistivity of the silicon or the conductance of an array of transistors. >From the haziness of my memory I'd say I'd read this between 2 and 1 years ago, or so. Rethinking about it now, I'm not sure how they avoided duping the resistivities or gains. The USG (LLNL?) has a patent (?) on conformal coatings containing lots of little reflective flecks embedded in a clear paint. Some kind of optical scan of the paint is reproducable but damn hard to duplicate. Similar to your turbulent deposition scheme. Good for keeping track of your nuclear inventory, etc.
