Modelagem e análise temporal da rede de transações de uma plataforma de consenso distribuído

• Autor(a) principal: Mascarenhas, Juliana Zanelatto Gavião
• Data de Publicação: 2019
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações do LNCC
• Texto Completo: https://tede.lncc.br/handle/tede/299
• topic: Análise de sistemas; Blockchains (Databases); Teoria dos grafos; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::SISTEMAS DE COMPUTACAO

Distributed consensus platforms have witnessed steady growth in recent years. Blockchain is the enabling technology behind this growth, in particular, because of its feature that allows transaction storage without the involvement of third trusted parties. This feature and its broad applicability in many areas have been the core of the technology’s success. Particularly, the blockchain applications on financial systems provided more notoriety and popularity to the technology. Bitcoin, the first-mover blockchain-based platform for cryptocurrency, is the most significant example. Nevertheless, another more recent platform called Ethereum is gaining a significant share on the cryptocurrence scenario. Despite the increasing interest for the blockchain technology, there are few analytical studies on the subject. For the majority of papers, the famous Bitcoin is the main subject of study. The remaining platforms, like Ethereum, compose a small part of these scientific works, even if it represents a significant portion of the distributed consensus platfrom in use. Some important related work have Ethereum as their main object of study. However, there are no temporal modelling and analysis in its scenario. The temporal analysis of a transaction allows verifying changes, the behavior and profile of users of the platform. In this way, the object of study of this dissertation are the transactions of the Ethereum platform. This platform has a vast range of operations, combining the blockchain and smart contracts technologies. Besides that, currently Ethereum is the second most used platform among cryptocurrencies. This work performs the data modeling of the transactions of Ethereum using the mathematical representation MultiAspect Graphs (MAGs). Differently of networks with a cyclic time period, the transactions performed by users are continuous and do not have regular patterns. The transactions are unique and don’t repeat in time, characterizing a system with an acyclic period. By inserting the temporal aspect in the analysis we observe the variation in behavior and the dynamics of the platform. A compact analysis, which aggregates time instants, do not allow the visualization or quantification of such changes. Therefore, the goal of this study is to present, through the analysis of the modeled network, the behavior and evolution of the Ethereum transacton network through time. Therefore, we show that the temporal analysis aggregates important information in a study with that goal.