fyi: NSEC5: Provably Preventing DNSSEC Zone,Enumeration

perhaps of interest, this talk will be given tomorrow Tue 26-Aug-2014 in 
Gates Hall, Stanford Univ (link to paper near bottom)...


Subject: Tuesday, August 26 -- Moni Naor: Primary-Secondary-Resolvers
  Membership Proof Systems and their Applications to DNSSEC
From: Joe Zimmerman <[email protected]>
Date: Mon, 25 Aug 2014 09:55:27 -0700
To: [email protected]

   Primary-Secondary-Resolvers Membership Proof Systems and
                 their Applications to DNSSEC

                          Moni Naor

                   Tuesday, August 26, 2014
                        Talk at 4:15pm
                          Gates 463A

Abstract:

We consider Primary-Secondary-Resolver Membership Proof Systems (PSR for
short)
that enable a secondary to convince a resolver whether or not a given a
element
is in a set defined by the primary without revealing more information about
the set.  The main motivation is studying the problem of zone enumeration
in DNSSEC. DNSSEC is designed to prevent network attackers from tampering
with domain name system (DNS) messages. The cryptographic machinery used
in DNSSEC, however, also creates a new vulnerability - Zone Enumeration,
where an adversary launches a small number of online DNSSEC queries and then
uses offline dictionary attacks to learn which domain names are present or
absent in a DNS zone.

We explain why current DNSSEC (NSEC3) suffers from the problem of zone
enumeration: we use cryptographic lower bounds to prove that in a PSR system
the secondary must perform non trivial online computation. This implies that
the three design goals of DNSSEC --- high performance, security against
network attackers, and privacy against zone enumeration --- cannot be
satisfied simultaneously.

We provide PSR constructions matching our lower bound and in particular
suggest NSEC5, a protocol that solves the problem of DNSSEC zone enumeration
while remaining faithful to the operational realities of DNSSEC. The scheme
can be seen as a variant of NSEC3, where the hash function is replaced with
an RSA based hashing scheme. Other constructions we have are based on the
Boneh–Lynn–Shacham signature scheme, Verifiable Random and Unpredictable
Functions and Hierarchical Identity Based Encryption.

The talk is based on the papers "NSEC5: Provably Preventing DNSSEC Zone
Enumeration" by Sharon Goldberg, Moni Naor, Dimitrios Papadopoulos, Leonid
Reyzin, Sachin Vasant and Asaf Ziv

   https://www.cs.bu.edu/~goldbe/papers/nsec5.pdf

  and "PSR Membership Proof Systems" by
Moni Naor and Asaf Ziv


_______________________________________________
dnsext mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dnsext