and for a whole lot of drift with respect to smartcards
being pda/cellphone wanabees
Storm building over RFID-enabled passports
http://www.networkworld.com/news/2006/092106-rfi
d-passports.html
from above:
The chip, which is embedded inside the cover of the
passport, contains only a duplicate copy of the passport
photograph and the printed data. The digital data is
intended to prevent forgeries by allowing inspectors to
compare the printed and digital data.
... snip ...
the article mentions that integrity of the electronic data
is protected by a digital signature (preventing tampering
and/or forgeries).
At some level, the digitally signed data can be considered a
electronic credential that is extremely difficult to
counterfeit.
posting with number of references about cloning (electronic)
passport data
http://www
.garlic.com/~lynn/aadsm25.htm#11 And another cloning
tale
from three factor authentication model
ht
tp://www.garlic.com/~lynn/subpubkey.html#3factor
* something you have
* something you know
* something you are
... frequently hardware tokens (chips) are implemented as
"something you have" authentication (i.e. the chip
supposedly contains some unique information ... which
differentiates it from every other chip). some recent posts
mentioning "something you have" authentication.
http://www
.garlic.com/~lynn/aadsm25.htm#30 On-card displays
http://www
.garlic.com/~lynn/aadsm25.htm#25 RSA SecurID SID800
Token vulnerable by design
http://www
.garlic.com/~lynn/aadsm25.htm#16 Fraudwatch -
Chip&PIN one-sided story
however, taking the passport chip data as an electronic
credential, cloning the information doesn't (directly)
represent a vulnerability ... since it is more analogous to
digital certificates ... which are readily assumed to be
widely distributable.
the passport chip data as an electronic credential
containing a digital photograph ... and matching a person's
face to the digital photograph then represents
"something you are" authentication (as opposed to
assuming the chip ...or even a cloned chip ... represents
any sort of "something you have" authentication).
in theory, an electronic credential would be considered
valid, regardless of any specific chip container that it
might be carried in. one might then make the assertion, that
a passport electronic
credential could be carried in any device capable of
reliably reproducing the correct bits.
going back to the issue raised in
http://www
.garlic.com/~lynn/aadsm25.htm#30 On-card displays
that most smartcards/chips are really pda/cellphone wanabees
... one might suggest that you could then even carry your
electronic credential/passport in your pda or cellphone ...
as opposed to needing a separate physical device.
the issue that then is raised are there any significant
privacy considerations similar to privacy issues raised with
x.509 identity digital certificates from the early 90s
(having large amounts of privacy information in x.509
identity digital certificates widely distributed all over
the place).
by the mid-90s, many institutions considered that the
privacy and liability problems with x.509 identity digital
certificates were so significant that they retrenched to
"relaying-party-only" certificates. lots of past
posts mentioning rpo-certificates
http:/
/www.garlic.com/~lynn/subpubkey.html#rpo
these were digital certificates that effectively only
contained some sort of database index or account number. the
relying party then used the account number to retrieve the
actual information of interest (w/o having to widely expose
it in any way).
the analogy for an electronic passport infrastructure would
be just needing to present the passport number. the actual
credential data (and any photos or other information
necessary for "something you are" authentication)
is retrieved from secure online repository.
as repeatedly pointed out in the "RPO" digital
certificate scenario ... it isn't even necessary to include
the account/passport number in a digitally signed document
... since there is no information that needs integrity
protection. the person just makes an assertion as to their
correct account/passport number. the appropriate information
is then retrieved from the online infrastructure and used
for authentication (and whatever other required purposes).
asserting the
wrong account/passport number presumably retrieves
information that fails to result in valid authentication.
needing (some certification authority) to digitally sign the
passport/account number (in the RPO scenario) for any
possible integrity purposes, is then redundant and
superfluous (one of my oft
repeated comments).
------------------------------------------------------------
---------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography"
to majordomo metzdowd.com
|
Doesn't
of course address the requirement that we want
evidence (such as a signed paper receipt) that can later be
adjudicated
by a court with higher evidential standards than a bank
statement that
their systems work perfectly...
- --
Blogzilla -->