论文标题
利用交易并发以加快区块链的速度
On Exploiting Transaction Concurrency To Speed Up Blockchains
论文作者
论文摘要
共识协议当前是防止区块链系统扩展的瓶颈。但是，我们认为交易执行对于区块链的性能和安全性也很重要。换句话说，有足够的机会通过降低交易执行成本来加快加快区块链的速度。 我们的目标是通过利用区块链工作负载中可用的交易并发性来了解我们可以加快区块链的速度。为此，我们首先分析了七个主要公共区块链的历史数据，即比特币，比特币现金，Litecoin，Dogecoin，Dogecoin，以太坊，以太坊经典和Zilliqa。我们考虑并发两个指标，即每个块的单交易冲突率，以及每个块的组冲突率。我们发现，基于UTXO的区块链的并发性比基于帐户的区块链有更多的并发性，尽管前者的并发量低于预期。另一个有趣的发现是，一些具有较大区块的区块链比具有较小区块的区块链具有更多的并发性。接下来，我们提出了一个分析模型，用于估计给定数量并发量的交易执行速度。使用我们的经验分析结果，该模型估计，如果利用所有可用并发性，可以实现以太坊中的6倍加速。
Consensus protocols are currently the bottlenecks that prevent blockchain systems from scaling. However, we argue that transaction execution is also important to the performance and security of blockchains. In other words, there are ample opportunities to speed up and further secure blockchains by reducing the cost of transaction execution. Our goal is to understand how much we can speed up blockchains by exploiting transaction concurrency available in blockchain workloads. To this end, we first analyze historical data of seven major public blockchains, namely Bitcoin, Bitcoin Cash, Litecoin, Dogecoin, Ethereum, Ethereum Classic, and Zilliqa. We consider two metrics for concurrency, namely the single-transaction conflict rate per block, and the group conflict rate per block. We find that there is more concurrency in UTXO-based blockchains than in account-based ones, although the amount of concurrency in the former is lower than expected. Another interesting finding is that some blockchains with larger blocks have more concurrency than blockchains with smaller blocks. Next, we propose an analytical model for estimating the transaction execution speed-up given an amount of concurrency. Using results from our empirical analysis, the model estimates that 6x speed-ups in Ethereum can be achieved if all available concurrency is exploited.
