Despite the great success of blockchain systems in recent years, blockchains still struggle to provide the same level of latency and throughput as centralized financial systems. The core of this problem lies in the inefficiency of consensus protocols. In this paper, we provide a survey on recent efforts to improve the scalability of blockchains. We focus on layer-two protocols, such as payment channel networks and transaction rollups, which process computations off-chain and only use consensus for dispute resolution. Layer-two protocols are expected to process microtransactions with sub-second latency and reduced fees, allowing blockchains to scale. Much of this work addresses the open challenges of payment channel networks, such as payment routing, channel rebalancing, network design strategies, security and privacy, payment scheduling, congestion control, simulators, and support for light nodes. We also dedicate a section to the existing implementations of smart-contract-based transaction rollups. Our work systematizes the state-of-the-art layer-two protocols, paving the way for future advances.

中文翻译：

---

区块链可扩展性调查：从硬件到第二层协议


尽管区块链系统近年来取得了巨大成功，但区块链仍然难以提供与中心化金融系统相同水平的延迟和吞吐量。这个问题的核心在于共识协议的低效率。在本文中，我们对最近为提高区块链可扩展性所做的努力进行了调查。我们专注于第二层协议，例如支付通道网络和交易汇总，它们在链外处理计算，并且仅使用共识来解决争议。第二层协议预计将以亚秒级延迟处理微交易并降低费用，从而使区块链能够扩展。这项工作的大部分工作解决了支付通道网络的开放挑战，例如支付路由、通道再平衡、网络设计策略、安全和隐私、支付调度、拥塞控制、模拟器以及对轻节点的支持。我们还专门用一节介绍基于智能合约的交易汇总的现有实现。我们的工作将最先进的第二层协议系统化，为未来的进步铺平了道路。