Ethereum is a new blockchain-based platform that is also capable of running smart contracts. Despite its increasing popularity, there is a lack of studies on characterizing this system, in special the fees paid by users and the respective delay to confirm the transactions, that is, the pending time. In this sense, we study the main features of Ethereum transactions and evaluate the common belief-for blockchain systems that rely on proof of work-that users who pay higher fees will have their transactions confirmed faster. Specifically, we collect information about 7.2 million of transactions in Ethereum and correlate their pending time to several fee-related features. Moreover, we conduct our study evaluating different ranges of values for the features, such as default and unusual values adopted by users as well as clusters of users with similar behaviors. Our empirical analysis shows strong evidence that there is no clear correlation between fees-related features and the pending time. Overall, we conclude from our investigation that transaction's features, including gas and gas price defined by users, cannot determine the pending time of transactions. | INTRODUCTIONWe have witnessed tremendous growth in blockchain-based platforms in recent years. In this context, Bitcoin 1 is the first and most notable blockchain-based platform, and cryptocurrency is still the most common application of the technology. However, since the ascension of Bitcoin, many other blockchain-based platforms emerged. For instance, Ethereum 2 has recently emerged as a popular cryptocurrency by reaching a market of more than 27 billion US dollars and with more than 470 million transactions * .Differently from Bitcoin, the Ethereum platform focuses on Smart Contracts to support financial transactions and deploy distributed applications. Smart contracts are autonomous computer programs that execute predefined logic when they are triggered. Ethereum, thus, provides more flexibility to its users and can achieve more complex results and fulfilling many Bitcoin gaps.Similarly to most of all other cryptocurrency systems, the Ethereum's central operation is the maintenance of a global and public ledger, known as the blockchain. The blockchain records all transactions between participants. Participants of a transaction sign it using their public-private keys. Then, this transaction must be added to a block, which *https://etherscan.io

Ethereum is one of the most popular cryptocurrency currently and it has been facing security threats and attacks. As a consequence, Ethereum users may experience long periods to validate transactions. Despite the maintenance on the Ethereum mechanisms, there are still indications that it remains susceptible to a sort of attacks. In this work, we analyze the Ethereum network behavior during an under-priced DoS attack, where malicious users try to perform denial-of-service attacks that exploit flaws in the fee mechanism of this cryptocurrency. We propose the application of machine learning techniques and ensemble methods to detect this attack, using the available transaction attributes. The proposals present notable performance as the Decision Tree models, with AUC-ROC, F-score and recall larger than 0.94, 0.82, and 0.98, respectively.

No abstract

Ethereum has emerged as one of the most important cryptocurrencies in terms of the number of transactions. Given the recent growth of Ethereum, the cryptocurrency community and researchers are interested in understanding the Ethereum transactions behavior. In this work, we investigate a key aspect of Ethereum: the prediction of a transaction confirmation or failure based on its features. This is a challenging issue due to the small, but still relevant, fraction of failures in millions of recorded transactions and the complexity of the distributed mechanism to execute transactions in Ethereum. To conduct this investigation, we train machine learning models for this prediction, taking into consideration carefully balanced sets of confirmed and failed transactions. The results show high-performance models for classification of transactions with the best values of F1-score and area under the ROC curve approximately equal to 0.67 and 0.87, respectively. Also, we identified the gas used as the most relevant feature for the prediction.

No abstract