Secure and efficient distributed/decentralized storage and computing

Point of contact

  • Benjamin Smith, INRIA-Saclay
  • Maryline Laurent, SAMOVAR-Telecom SudParis

Summary

Distributed/decentralized systems can advantageously support properties of interest, such as high availability, and distributed governance in decision-making and computing. One of the key technologies underlying this is blockchain, on which a number of services are currently emerging, including traceability, authentication, and identity management.

The main challenges with regard to these systems are related to making distributed/decentralized storage and computing related services as efficient, reliable and flexible as centralized services. Investigations focus on privacy-enhancing technologies, such as secure multiparty computation where efforts are still needed to reduce computation and communication overhead; homomorphic encryption, which still requires complexity reduction for large adoption in cloud environments; and searchable encryption which is still on demand for query expressiveness, verifiability and accountability of the search results.

Another challenge is to extend the threat model range. While the usual threat models consider the honest-but-curious adversary where entities run the protocol correctly but attempt to collect data, there is a need to decrease security assumptions, moving towards malicious adversary models where adversaries are allowed to not run the protocol properly. This requires complementary proof mechanisms being designed and integrated into already complex schemes.

Keywords

secure multiparty computation, homomorphic encryption, searchable encryption

Researchers involved or interested

  • Maryline Laurent, Samovar
  • Joaquin Garcia-Alfaro, Samovar
  • Nesrine Kaâniche, Samovar
  • Françoise Levy-dit-Vehel, GRACE