From network-wide to localized: some consensus but not too much!
Speaker: Davide Frey
INRIA Rennes
DEIB - Seminar Room "A. Alario" (Bld. 21)
October 11th, 2024 | 10.30 am
Contact: Prof. Alessandro Margara
INRIA Rennes
DEIB - Seminar Room "A. Alario" (Bld. 21)
October 11th, 2024 | 10.30 am
Contact: Prof. Alessandro Margara
Abstract
On October 11th, 2024 at 10.30 am the seminar titled "From network-wide to localized: some consensus but not too much!" will take place at DEIB Seminar Room "Alessandra Alario" (Building 21).
The advent of Bitcoin in 2008 brought renewed interest in decentralized applications.
Albeit initially devoted to cryptocurrency, the blockchain model has found applications in a number of settings such as logistics, self-sovereign identity systems, or e-voting. Today, however, we know that the main blockchain application, cryptocurrency, does not need the full power provided by the blockchain abstraction. In particular, it does not require the total order of operations provided by consensus. However, the question remained open regarding other blockchain applications. In this talk consider three of them: anonymous cryptocurrency, self-sovereign identity, and decentralized e-voting. We show that, unlike plain cryptocurrency, their implementation does require consensus. However, contrary to general practice, this consensus does not need to involve the entire network. The talk introduces two novel distributed objects: the AllowList and the DenyList. It shows that the AllowList can be implemented without consensus, while the DenyList only needs a restrained form of consensus involving a small set of participants. The talk then shows that AllowList and DenyList can implement the three considered applications. Finally, it introduces CAC, a novel primitive that generalizes the concept of restrained consensus.
Davide Frey has been a researcher at Inria Rennes Bretagne-Atlantique since 2010. He received his PhD from Politecnico di Milano in 2006 and worked as a post-doc at Washington University in St. Louis and Inria Rennes. His research focuses on distributed systems and algorithms, with contributions in content dissemination, social networks, blockchain, and decentralized recommendation and machine-learning systems. He has also worked on combinatorial optimization and lightweight blockchain alternatives. He is active in national and international projects, including the European H2020 SOTERIA project and the French ANR Byblos project on Beyond-blockchain data structures. He also coordinates the PriCLeSS Cominlabs project, collaborating with computer scientists and law researchers on privacy and legal issues of blockchain-like infrastructures.
The advent of Bitcoin in 2008 brought renewed interest in decentralized applications.
Albeit initially devoted to cryptocurrency, the blockchain model has found applications in a number of settings such as logistics, self-sovereign identity systems, or e-voting. Today, however, we know that the main blockchain application, cryptocurrency, does not need the full power provided by the blockchain abstraction. In particular, it does not require the total order of operations provided by consensus. However, the question remained open regarding other blockchain applications. In this talk consider three of them: anonymous cryptocurrency, self-sovereign identity, and decentralized e-voting. We show that, unlike plain cryptocurrency, their implementation does require consensus. However, contrary to general practice, this consensus does not need to involve the entire network. The talk introduces two novel distributed objects: the AllowList and the DenyList. It shows that the AllowList can be implemented without consensus, while the DenyList only needs a restrained form of consensus involving a small set of participants. The talk then shows that AllowList and DenyList can implement the three considered applications. Finally, it introduces CAC, a novel primitive that generalizes the concept of restrained consensus.
Davide Frey has been a researcher at Inria Rennes Bretagne-Atlantique since 2010. He received his PhD from Politecnico di Milano in 2006 and worked as a post-doc at Washington University in St. Louis and Inria Rennes. His research focuses on distributed systems and algorithms, with contributions in content dissemination, social networks, blockchain, and decentralized recommendation and machine-learning systems. He has also worked on combinatorial optimization and lightweight blockchain alternatives. He is active in national and international projects, including the European H2020 SOTERIA project and the French ANR Byblos project on Beyond-blockchain data structures. He also coordinates the PriCLeSS Cominlabs project, collaborating with computer scientists and law researchers on privacy and legal issues of blockchain-like infrastructures.