On the transaction dynamics of the Ethereum-based cryptocurrency

Abstract Distributed blockchain-based consensus platforms have witnessed steady growth in recent years. In special, cryptocurrency is one of the main applications of the blockchain technology. Despite the recent interest in blockchain, we still lack in-depth analysis of systems that use such a technology. In fact, most of the existing works focus on Bitcoin. Moreover, blockchain-based cryptocurrency systems are highly dynamic. Their internal mechanisms and consensus algorithms evolve over time. Users also change their interests in a given platform, which in turn, reflect their behaviour. In this article, we model the Ethereum-based cryptocurrency transaction network, a more recent blockchain platform that is gaining a significant share in the cryptocurrency market. We model the transactions of Ethereum as a complex system, representing this complex system as a time-varying graph. Our model and the analysis we conduct rely on a 3-year dataset of Ethereum-based cryptocurrency transactions, comprising more than 38 million users (i.e. unique wallet addresses) and almost 300 million transactions. We analyse the evolution of users and transactions over time. Our study also highlights the centralization tendency of the transaction network on both user and time aspects. Finally, we also analyse the formation of communities and the evolution of connected components considering the dynamics of the Ethereum-based cryptocurrency transaction network.