IN3’s K-ryptography and Information Security for Open Networks research group is pleased to invite you to the Seminar «Blockchain, Cryptography and Machine Learning» with the participation of Adrià Torralba-Agell, PhD Candidate at KISON research group, and Ghazaleh Keshavarzkalhori, researcher of the SENDA research group at the Universitat Autònoma de Barcelona.
The seminar will be held on Tuesday 11 June at 18:00 h (CET) at Interdisciplinary R&I Hub (Building C - Room C0.6).
Venue
Interdisciplinary R&I Hub (Building C - Room 101 (C0.6))
Rambla del Poblenou, 154
08018 Barcelona
Espanya
When
11/06/2024 18.00h
Organized by
Universitat Oberta de Catalunya, IN3's K-ryptography and Information Security for Open Networks research group
Program
Summary
Seminar «A Taxonomy of Security Analysis of Blockchain Layer 2 Scalability Solutions and Future Directions» - Adrià Torralba-Agell
Blockchains, heralded for its transformative impact across various sectors, confronts inherent scalability challenges that impede widespread adoption. Layer 2 solutions emerge as a promising avenue to alleviate these limitations, offering scalability enhancements while trying to retain the core security guarantees of underlying blockchains. However, despite their increasing prominence, the security assumptions underpinning layer 2 solutions have been insufficiently scrutinized. Addressing this gap, this paper contributes a rigorous examination of the security assumptions inherent in the most used blockchain layer 2 scalability solutions (Payment Channel Networks, Optimistic Rollups, Zero-Knowledge Rollups, Validium and Volition). Our analysis reveals that layer 2 solutions need additional security assumptions beyond those already established by their underlying blockchains (layer 1). By delineating these assumptions and their implications, our research facilitates a deeper understanding of the security challenges inherent in blockchain scalability solutions.
Adrià Torralba-Agell
PhD Candidate at Universitat Oberta de Catalunya (UOC) in the K-ryptography and Information Security for Open Networks (KISON) Research Group. He received the double BSc in Mathematics and Compute Science from Universitat de Barcelona (UB), as well as, a MSc on Fundamental Principles of Data Science. His research interests include Abstract Algebra, Zero-Knowledge Cryptography, Interactive Proof Systems, Probabilistic Proofs, and Blockchain Security.
Seminar «A Verifiable Federated Learning Scheme based on zkSNARKs and Blockchain» - Ghazaleh Keshavarzkalhori
With the distribution of computations, verifiability and retraction of trust is become important. One widely used distribution computation setting is distributed machine learning, where a model isn’t trained by ourselves and by different untrusted servers. In such a setting, verifying the training process and the computations of remote servers is important. Federify is a work designed to bring verifiability to such a setting with the help of blockchains and snarks.
Ghazaleh Keshavarzkalhori
B.S. in Electrical Engineering from Tehran University, specializing in communications. She has mainly researched in the fields of blockchain technology and privacy preserving protocols at the SCL of Tehran University. After graduating, she has worked as a researcher in the SENDA research group at the Universitat Autònoma de Barcelona. She is researching the combination of federated learning, blockchain and privacy.