Efficient private database queries using ring-LWE somewhat homomorphic encryption

Tushar Kanti Saha*, Mayank Rathee, Takeshi Koshiba

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)


We study the problem of private database queries over any outsourced encrypted database. In 2016, Kim et al. [IEEE Trans. on Dependable and Secure Comput.] showed three private query processing protocols for conjunctive, disjunctive, and threshold queries respectively over an encrypted database. First, we propose two more efficient protocols of processing conjunctive and disjunctive queries with a lower-depth equality circuit than Kim et al.’s one. To get the lower-depth circuit, we modify the packing methods of Saha and Koshiba [APWConCSE 2016] to support an efficient batch computation of our protocols within a few multiplications using binary encoded data. Secondly, we propose another efficient protocol engaging a batch technique for threshold query processing in which many inequality comparisons are indispensable. Finally, we propose the packing methods to support batch computation of many inequality circuits. In addition, our theoretical analysis along with practical experiments show that our protocols are more efficient than any other existing method. We further enhance the performance of conjunctive and disjunctive query protocols using a base-N fixed-length encoding.

Original languageEnglish
Article number102406
JournalJournal of Information Security and Applications
Publication statusPublished - 2019 Dec


  • Batch technique
  • Conjunctive queries
  • Disjunctive queries
  • Homomorphic encryption
  • Packing method
  • Private database queries
  • Threshold queries

ASJC Scopus subject areas

  • Software
  • Safety, Risk, Reliability and Quality
  • Computer Networks and Communications


Dive into the research topics of 'Efficient private database queries using ring-LWE somewhat homomorphic encryption'. Together they form a unique fingerprint.

Cite this