Swanand Kadhe
Department of Electrical Engineering and Computer Sciences
University of California, Berkeley

Abstract
The emerging field of blockchains and distributed ledgers has far-reaching applications such as healthcare, supply-chain, and Internet-of-Things. However, one of the main roadblocks in the proliferation of blockchain applications is that current blockchain designs are resource-intensive. In this talk, we will focus on blockchain storage requirements that are growing near-exponentially. We propose an architecture based on 'fountain codes', a class of erasure codes, that enables any user to 'encode' validated blocks into a small number of 'coded blocks', thereby reducing their storage costs by orders of magnitude. In particular, our proposed "Secure Fountain (SeF)" architecture can achieve a near-optimal trade-off between the storage savings per user and the 'bootstrap cost' in terms of the number of (honest) storage-constrained users that a new user needs to contact to recover the blockchain. We demonstrate that the peeling decoder admitted by fountain codes turns out to be crucial for security against adversarial users that can provide maliciously formed coded blocks. Further, the rateless property of fountain codes helps in achieving high decentralization and scalability. Our experiments demonstrate that SeFcodes tuned to achieve 1000x storage savings enable users to encode the 191GB Bitcoin blockchain into 195MB on average, at the expense of 5% increase in the required bootstrap cost. This is a joint work with JichanChung and Kannan Ramchandran.

Bio
Swanand Kadheis a postdoctoral researcher in the EECS Department at UC Berkeley. He received his Ph.D. degree in electrical engineering from Texas A&M University in2017. He is a recipient of the 2016 Graduate Teaching Fellowship from TAMU College of Engineering. His research interests are in coding theory, information theory, and distributed computing, with applications to privacy and security in machine learning and blockchains.

Host
Professor Chih-Chun Wang, chihw@purdue.edu, 49-45568