Efficient scalable multiparty private set-intersection via garbled bloom filters

Roi Inbar, Eran Omri, Benny Pinkas

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

34 Scopus citations


In private set intersection (PSI), a set of parties, each holding a private data set, wish to compute the intersection over all data sets in a manner that guarantees both correctness and privacy. This secure computation task is of great importance and usability in many different real-life scenarios. Much research was dedicated to the construction of PSI-tailored concretely efficient protocols for the case of two-party PSI. The case of many parties has been given much less attention, despite probably being a more realistic setting for most applications. In this work, we propose a new concretely efficient, highly scalable, secure computation protocol for multiparty PSI. Our protocol is an extension of the two-party PSI protocol of Dong et al. [ACM CCS’13] and uses the garbled Bloom filter primitive introduced therein. There are two main variants to our protocol. The first construction provides semi-honest security. The second construction provides (the slightly weaker) augmented semi-honest security, and is substantially more efficient. Furthermore, in the augmented semi-honest protocol all heavy computations can be performed ahead of time, in an offline phase, before the parties ever learn their inputs. This results in an online phase that requires only short interaction. Moreover, in the online phase, interactions are performed over a star topology network. All our constructions tolerate any number of corruptions. We implemented our protocols and incorporated several optimization techniques. These techniques allow the running time of the protocol to be comparable to that of the two party protocol of Dong et al. and scale linearly with the number of parties. We ran extensive experiments to compare our protocol with the two-party protocol and to demonstrate the effect of the different optimizations.

Original languageEnglish
Title of host publicationSecurity and Cryptography for Networks - 11th International Conference, SCN 2018, Proceedings
EditorsDario Catalano, Roberto De Prisco
PublisherSpringer Verlag
Number of pages18
ISBN (Print)9783319981123
StatePublished - 2018
Event11th International Conference on Security and Cryptography for Networks, SCN 2018 - Amalfi, Italy
Duration: 5 Sep 20187 Sep 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11035 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference11th International Conference on Security and Cryptography for Networks, SCN 2018

Bibliographical note

Funding Information:
R. Inbar—Research supported by ISF grant 544/13. E. Omri—Research supported by ISF grants 544/13 and 152/17. B. Pinkas—Research supported by the BIU Center for Research in Applied Cryptography and Cyber Security in conjunction with the Israel National Cyber Bureau in the Prime Minsters Office, and by ISF grant 1018/16.

Publisher Copyright:
© 2018, Springer Nature Switzerland AG.


  • Concrete efficiency
  • Garbled Bloom filters
  • Multiparty computation
  • Private set intersection


Dive into the research topics of 'Efficient scalable multiparty private set-intersection via garbled bloom filters'. Together they form a unique fingerprint.

Cite this