Measurement-based analyses of public blockchain networks

Abstract

The Bitcoin network has been online for a decade to successfully prove that the blockchain is a promising technology for the peer-to-peer electronic cash system. Numerous public blockchain systems have come after not only to achieve better scalability but also to expand blockchain applications to various industries. Ethereum, being most successful among them, has envisaged a blockchain system with smart contracts for decentralized applications. Due to the successful operation of decentralized blockchain networks for several years, Bitcoin and Ethereum public blockchains have accumulated billions of transactions arranged in millions of blocks. Accordingly, historical block and transaction data have become a popular source of studying eco- nomic aspects of cryptocurrencies such as cryptocurrency abuse, anti-money laundering, and monetary flow of businesses. Block explorer services have emerged to provide transaction information stored in those blockchains with additional analytics and user-friendly interfaces. However, the canonical blockchain data is limited to provide block and transaction records in a historical view, i.e., the eventual ordered set of transaction records, that are only the final product of the consensus mechanism among distributed blockchain network nodes. Live block and transaction records corresponded by the network nodes are more abundant and variable, so that observing them plays an important role in understanding the systematic aspects of public blockchain systems. In this dissertation, we aim to broaden our understanding of blockchain systems through public blockchain measurement. Having established a sound and comprehensive measurement of the network, we conduct in-depth analyses in terms of the discrepancies and the characteristics of the observed events. Our analysis results not only unveil unusual behaviors of blockchain miners and users, but also provide the compelling reasons behind them. The dissertation further aims to enhance the security of blockchain nodes in a peer-to-peer networking perspective by presenting a new blockchain partitioning attack and its countermeasures, which are motivated by our measurement findings.