Secure Spontaneous Device Association (TECHNOTE)
Tim Kindberg and Kan Zhang
Hewlett-Packard Laboratories
one to many assoc btw people and devices
making spontaneous device interaction secure by bootstrapping from the physical properties of lasers
ex: sending photos wirelessly to a projector -- but only THAT projector
ex: mobile game play with lots of people -- picking teams
fixed locations vs. mobile devices
- trust
-- I trust some devices and people, but not others
-- some locations are more [intimate] than others
-- may be established in an ad hoc way
constraints
- no "logging in" -- no pre-established relationship
- no trusted third-party
virtual wire
- want effect of wire (unambiguous endpoints, secrecy, integrity)
- want physically validated key exchange w/o cable
physically constrained channels
- contact inconvenient
- line-of-sight inaccurate and hard to engineer
one-way key exchange
- one-way validated binding of key <--> device
- send wireless network address and secret key over laser beam
- other device responds - and then the first device checks it.
- physical counterpart to crypto-based authentication
two-way validated binding of key <--> device
- so then a light only goes on if the exchange goes correctly
- or a symmetric protocol
-- each party points a laser beam at the other
- difficult to intercept laser
- unlikely with 2-way, symmetric protocol b/c two humans have to physically point at each other
- evidence as bootstrap: physical characteristics perceived by humans
user investigation
- perceptions and reasoning about threats and trust
- help users behave in a secure way
various types of physically constrained channel
- compare evidential value
- de facto system properties
Q: Christa Monsch UCI Irvine
Why lasers?
A: b/c it's visible and communicative
