Saturday, October 11, 2008
« LZfuz: a fast compression-based fuzzer f... | Main | Detect Rootkits »
Revoking Repeatedly Misbehaving Anonymous Users Without Relying on TTPs
BLAC: Revoking Repeatedly Misbehaving Anonymous Users Without Relying on TTPs

Dartmouth Technical Report TR2008-635

        Patrick P. Tsang
        Man Ho Au
        Apu Kapadia
        Sean W. Smith

Date: October 2008


Abstract:
  Several credential systems have been proposed in which users can
 authenticate to service providers anonymously. Since anonymity can
 give users the license to misbehave, some variants allow the selective
 deanonymization (or linking) of misbehaving users upon a complaint to
 a trusted third party (TTP). The ability of the TTP to revoke a user's
 privacy at any time, however, is too strong a punishment for
 misbehavior.  To limit the scope of deanonymization, systems have been
 proposed in which users are deanonymized if they authenticate ``too
 many times,'' such as ``double spending'' with electronic cash. While
 useful in some applications, it is not possible to generalize such
 techniques to more subjective definitions of misbehavior, e.g., it is
 not possible to block users who ``deface too many webpages'' on a
 website.
  We present BLAC, the first anonymous credential system in which
 service providers can revoke the credentials of repeatedly misbehaving
 users without relying on a TTP. Since revoked users remain
 anonymous, misbehaviors can be judged subjectively without users
 fearing arbitrary deanonymization by a TTP. Finally, our
 construction supports a $d$-strikes-out revocation policy,
 whereby users who have been subjectively judged to have repeatedly
 misbehaved at least $d$ times are revoked from the system.

Note:
  The preliminary part of this report is based on our work originally
 published in the proceedings of the 14th ACM Conference on Computer
 and Communication Security (CCS '07), October 2007, under the title of
 "Blacklistable Anonymous Credentials: Blocking Misbehaving Users
 without TTPs". At the same period of time, we published an extended
 version of the CCS '07 paper as Dartmouth Computer Science technical
 report TR2007-601, which contained further details that did not fit
 into the conference paper.
  In this report, we make a significant additional contribution by
 extending our original construction of BLAC to provide more flexible
 revocation --- SPs can specify a d-strikes-out revocation policy, so
 that users can authenticate anonymously only if they have not
 misbehaved d or more times. Such a policy forgives a few (i.e., up to
 d-1) misbehaviors, but then blocks users who misbehave
 repeatedly. Following authentication, users remain anonymous, and SPs
 learn only whether a user has crossed the threshold of d
 misbehaviors. The original construction of BLAC is a special case with
 d=1.

Network Newsletters
Saturday, October 11, 2008 11:19:46 PM (Eastern Daylight Time, UTC-04:00)    Disclaimer  |  Comments [0]  |  Related posts:
SITESEEING: www.change.gov
AT&T Monthly Bandwidth Caps Are Here
ICANN proposes new way to buy top-level domains - Network World
Detect Rootkits
LZfuz: a fast compression-based fuzzer for poorly documented protocols
FCC Begins to Resolve Mutually Exclusive Noncommercial FM Radio Applications