IACR News
Here you can see all recent updates to the IACR webpage. These updates are also available:
03 April 2018
Anat Paskin-Cherniavsky
In addition to the obvious theoretical appeal of the question towards better understanding secure computation, perfect, as opposed to statistical reductions may be useful for designing MPC protocols with high concrete efficiency, achieved by eliminating the dependence on a security parameter.
1-out-of-2 bit-OT (dubbed OT) was shown to be complete for statistically secure 2PC for all functionalities [Kil88, IPS08]. Existing protocols in the OT-hybrid model only offer statistically secure with abort (efficient) protocols (requiring no further computational assumptions). In general, fairness can not be guaranteed, and only security with abort is achievable [Cleve86]. If the protocol is not required to be efficient in the security parameter $k$, then all 2PC functionalities can be securely evaluated [GK10] with statistical security in the OT-hybrid model.
As opposed to the statistical setting, it is not known whether OT is complete for perfectly secure 2PC. Furthermore, only a few examples of functionalities that have such protocols are known: we are only aware of string-OT and TO (OT with reversed roles) as perfectly reducible to OT. On the negative side, a large class is known, as implied by the fairness literature. By definition, functionalities not securely computable with fairness require super-polynomial in $k$ computational (and round) complexity to evaluate with error $neg(k)$ in the OT-hybrid model. This implies that these functionalities not computable with fairness are also not computable with perfect security (in the OT-hybrid model). For symmetric boolean functionalities, this class been fully characterized [ABMO15].
Back to the statistical world, quite surprisingly [IKOPS11] demonstrate that all client-server functionalities can be efficiently reduced to OT with statistical full security (no abort) in only one round.
Motivated by this relative ``ease'' of client-server functionalities for statistically secure 2PC in the OT-hybrid model, we study perfect reductions to OT for this class of functions. We prove that for many client-server functions of the form $f: X\times Y\rightarrow \{0,1\}$, where server domain size $|Y|$ is larger than client domain size $X$, have a perfect reduction to OT. More precisely, a $g(|X|,|Y|)=\Omega(1)$-fraction of functions are perfectly reducible to OT. This fraction grows roughly as $1-exp(|X|-|Y|)$. Furthermore, our reduction is 1-round using an oracle to secure evaluation of ${\text{OT}}^l$ (as in [IKOPS11]). As an example, this class contains $\text{2-out-of-5-OT}$. More generally, for $f: X\times Y\rightarrow Z$, $\Omega(1)$ of the functions with $|Y|>|X|(|Z|-1)$ are perfectly reducible to OT in 1 round.
Our work leaves open the question of whether all finite client-server functionalities are perfectly reducible to OT (not necessarily in one round). Another open question is whether 2PC functionalities that do have perfectly secure protocols in the OT hybrid model differ in round complexity, as is the case for statistical protocols.
Travis Scholl
Chris Brzuska, Antoine Delignat-Lavaud, Konrad Kohbrok, Markulf Kohlweiss
Olivier Bernard, Renaud Dubois, Simon Masson
Peizhao Hu, Sherman S.M. Chow, Asma Aloufi
Bernardo David, Rafael Dowsley, Mario Larangeira
Rafael Pass, Elaine Shi
Estuardo Alpirez Bock, Chris Brzuska, Wil Michiels, Alexander Treff
02 April 2018
Pascal Mainini, Rolf Haenni
This is an extended version of a paper accepted and presented at the Voting18 workshop of the Financial Cryptography and Data Security 2018 conference. It will be included in the conferences LNCS proceedings and available on the Springer web site.
University of Tartu, Estonia
We expect candidates to be able to develop and devote significant time to their own research agenda around the theme of the project. Successful candidates will help to design and evaluate privacy-enhancing cryptographic techniques for blockchains (e.g., SNARKs) and perform other research duties to help with the project, collaborate with partners and ensure the smooth administration of the project including the timely delivery of research output.
The EU H2020 project PRIViLEDGE requires travel to and collaboration with colleagues throughout the European Union. Full travel and equipment budget is available to support the activities of the project.
For any inquiries or to apply for the positions, submit a full research curriculum-vitae (cv), names of two references, and a research statement to Prof Helger Lipmaa (firstname.lastname (at) ut.ee) clearly indicating the position sought. This is crucial since we have several open positions.
The project started from January 1, 2018, and will last for three years. In the case of interest, the candidates may later seek further employment but this is not necessarily guaranteed. The position will stay open until we find a suitable candidate; please apply early.
Closing date for applications: 1 May 2018
Contact: Helger Lipmaa
More information: https://crypto.cs.ut.ee/index.php/Projects/PRIViLEDGE
01 April 2018
Adelaide, Australia, 7 December - 8 December 2018
Submission deadline: 15 June 2018
Oriahovitza, Bulgaria, 8 July - 15 July 2018
Amsterdam, The Netherlands, 13 September 2018
Submission deadline: 18 June 2018
Notification: 20 July 2018
Montpellier, France, 12 November - 14 November 2018
Submission deadline: 13 July 2018
Notification: 14 September 2018
Oslo, Norway, 28 November - 30 November 2018
Submission deadline: 10 August 2018
Notification: 10 September 2018
30 March 2018
Stephen Farrell
29 March 2018
Singapore University of Technology and Design (SUTD)
I am looking for PhD interns with interest in cyber-physical system security (IoT, water, power grid, transportation, and autonomous vehicle etc.). The attachment will be at least 3 months. Allowance will be provided for local expenses.
Interested candidates please send your CV with a research statement to Prof. Jianying Zhou.
Closing date for applications: 30 May 2018
Contact: Prof. Jianying Zhou
Email: jianying_Zhou (at) sutd.edu.sg
More information: http://jianying.space/
Luke Valenta, Nick Sullivan, Antonio Sanso, Nadia Heninger
Matteo Campanelli, Rosario Gennaro
Bertram Poettering, Paul Rösler
In this article we resolve the limitations of prior work by developing alternative security definitions, for unidirectional RKE as well as for RKE where both parties can contribute. We follow a purist approach, aiming at finding strong yet convincing notions that cover a realistic communication model; in particular, and in contrast to prior work, our models support fully concurrent operation of both participants.
We further propose secure instantiations (as the protocols analyzed or proposed by Cohn-Gordon et al. and Bellare et al. turn out to be weak in our models). While our scheme for the unidirectional case builds on a generic KEM as the main building block (differently to prior work that requires explicitly Diffie-Hellman), our schemes for bidirectional communication require, perhaps surprisingly, considerably stronger tools.