Loading...

Analysis and Improvement of Private Set Intersection Schemes

Kavousi, Alireza | 2020

574 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 53585 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Mohajeri, Javad; Salmasizadeh, Mahmoud
  7. Abstract:
  8. Secure multi-party computation (MPC) enables a group of mutually distrustful parties to compute a joint and agreed upon function of their private inputs without disclosing anything but the corresponding output. One of the most important secure computation protocols is private set intersection (PSI). In PSI, often two or several parties wish to find the intersection of their sets without revealing other non-common elements. There exist some other variants of PSI protocol like PSI cardinality or threshold PSI which in the former only the cardinality of the intersection set is revealed and in the latter the intersection set is revealed if its cardinality is greater (less) than a certain value. During these years, researchers have been on the quest to find more efficient protocols for typical PSI functionality or other variants of it. In this thesis, at the first stage we plan to study some necessary tools and concepts for analyzing PSI schemes. Then, we introduce some existing PSI protocols and discuss their properties and, in the end, we propose some improvements for existing protocols as well as presenting efficient two-party and multi-party PSI protocols. In detail, our innovations can be divided into three sections. In the first part, we show an existing PSI scheme is vulnerable against eavesdropping attack and then present a modified scheme which is secure against passive attacks without needing to have any secure channels and has less computational complexity compared to the previous scheme. In the second part, after introducing a two-party PSI protocol we show how to extend the scheme to an updatable multi-party PSI protocol using outsourced computation. In the last part, we present an efficient two-party PSI protocol which regarding the size hiding feature of at least one of the parties has less communication complexity compared to the protocol proposed by Chase and Miao in CRYPTO 2020. In addition, by extending the proposed scheme we present a multi-party PSI protocol which has less overhead compared to the state-of-the-art multi-party schemes
  9. Keywords:
  10. Private Set Intersections ; Secure Multiparty Computation (SMC) ; Oblivious Transfer Protocol ; Computational Complexity ; Communication Complexity

 Digital Object List

 Bookmark

No TOC