[Issuedrinks] [Fwd: Re: Fwd: rfid]

Koen Martens gmc at sonologic.nl
Wed Aug 9 21:20:30 UTC 2006

Sprak (mailde..) er met iemand die ik ken (Harald Welte, aka
LaForge, van ccc o.a.) over, en die had nog wel wat interresante
opmerkingen (geforward met permissie):

-------- Original Message --------

> I was talking to enno in irc today, about the rfid guardian
> (http://www.rfidguardian.org/).
> They don't seem to publish any schematics or anything, but i thought
> such an appliance would be actually quite simple to build. RFID is
> basically a binary search + some serial protocol, so i guess a
> simpel pic with some rudimentary send/receive circuit ought to be
> enough.

I disagree.  It's not at all "quite simple".  A number of factors you
(and I guess also the rfidguardian people to a certain degree) forget to
take into consideration:

1) RFID is not one standard/prtocol.  There are half a dozen different
   publicly documented standard (ISO 14443-1,2,3,4, ISO 15693,
   ISO18000-6, ... ) and a number of vendor-specific proprietary systems
   such as mifare, legic, etc. for which very little to no information
   is known.  some of those systems use binary search, but they all use
   different modulation / framing / clocking / frequency / etc.
2) Many commonly found applications (ePassport, electronic public
   transport ticketing, ... use encryption, so any man in the middle
   'monitoring and regulating use' will be impossible.
3) the magnetic coupling based systems have extremely short range, and
   the reader-transponder vs. transponder-reader signals are at least
   60dB apart (dynamic range problem).

> Anyway, enno thought you might already be working on something like
> that, so i thought i'd just send you an email. I'm interrested in
> stuff like this, that's my motivation :)

Milosch (see Cc) and I have just finished a free (GPL licensed firmware,
CC-sharealike-attribution schematics, ..) RFID reader design that can be
used for mainly ISO 14443 and 15693 but probably other 13.56MHz based
systems.  We're also working (contract based) on a 14443 PICC
(transponder-side) emulator, which will probably be released at some
point in september, also under GPL/CC licenses.

There are many issues that need to be looked into, stuff like
reconfigurable higher-order analoge filters, and that's only for
supporting all bitrates of only ISO 14443.  Our devices will run on a
Atmel AT91SAM7 (ARM7TDMI cpu) controller running at 48MHz clock, not a
small PIC ;)

So if you want to get into implementing any such device, please don't
make the error of underestimating the amount of work required.

And we're just talking about the 'usual' reader and card interface, not
about actively jamming signals (which would have to be bit clock and
phase accurate) or anything like that.  In theory this is all possible,
but actually designing a device that does it is a somewhat different
task.  Also, it's difficult to jam a _magnetically_ coupled signal over
distance, since the magnetic field very quickly deterioates with
distance (compared with the electric field).  But having a simulated tag
in the field (<10cm) of a 13.56MHz reader would certainly provide and
ideal opportunity to cause collisions with about every known anti
collision algorithm :)

... and once we're finished with the task for 13.56MHz, EPCglobal /
ISO18000-6 systems will start to be introduced in supermarkets,
providing new challenges at handling 860-920MHz signals with backscatter

- Harald Welte <laforge at gnumonks.org>          	
We all know Linux is great...it does infinite loops in 5 seconds. --

K.F.J. Martens, Sonologic, http://www.sonologic.nl/
Networking, hosting, embedded systems, unix, artificial intelligence.
Public PGP key: http://www.metro.cx/pubkey-gmc.asc
Wondering about the funny attachment your mail program
can't read? Visit http://www.openpgp.org/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: file:///home/gmc/tmp/nsmail.tmp
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://mailman.sonologic.nl/pipermail/issuedrinks/attachments/20060809/4b5da1b9/nsmail.bin

More information about the Issuedrinks mailing list