Agreement with Satoshi - On the Formalization of Nakamoto Consensus

The term Nakamoto consensus is generally used to refer to Bitcoin’s novel consensus mechanism, by which agreement on its underlying transaction ledger is reached. It is argued that this agreement protocol represents the core innovation behind Bitcoin, because it promises to facilitate the decentralization of trusted third parties. Specifically, Nakamoto consensus seeks to enable mutually distrusting entities with weak pseudonymous identities to reach eventual agreement while the set of participants may change over time. When the Bitcoin white paper was published in late 2008, it lacked a formal analysis of the protocol and the guarantees it claimed to provide. It would take the scientific community several years before first steps towards such a formalization of the Bitcoin protocol and Nakamoto consensus were presented. However, since then the number of works addressing this topic has grown substantially, providing many new and valuable insights. Herein, we present a coherent picture of advancements towards the formalization of Nakamoto consensus, as well as a contextualization in respect to previous research on the agreement problem and fault tolerant distributed computing. Thereby, we outline how Bitcoin’s consensus mechanism sets itself apart from previous approaches and where it can provide new impulses and directions to the scientific community. Understanding the core properties and characteristics of Nakamoto consensus is of key importance, not only for assessing the security and reliability of various blockchain systems that are based on the fundamentals of this scheme, but also for designing future systems that aim to fulfill comparable goals. I. THE IMPORTANCE OF NAKAMOTO CONSENSUS The recent explosive increase in both economic valuation and technical interest towards Bitcoin, cryptocurrencies, and distributed ledgers in general, is mirrored by equally growing research efforts from the scientific community to better understand, employ and extend upon the fundamental principles that govern these technologies. Not only has the body of peer-reviewed literature directly related to Bitcoin and cryptocurrencies increased substantially as outlined by this recent taxonomy [XWS17], but a lot of work is also presented online in the form of pre-prints, e-prints, and informally as blog posts, following the publishing spirit of the original Bitcoin white paper [Nak08a]. Moreover, many other research fields are also exploring how the underlying concepts behind blockchain technologies could be applied in their domain. This renders it difficult for both researchers and practitioners to get a coherent picture of the state-of-the-art in this emerging field. We therefore 1One of the earliest uses of the term Nakamoto consensus can be attributed to a blog post by Nick Szabo in [Sza14], after which it appeared in scientific publications such as [BMC+15], [LTKS15]. believe that further systematization efforts related to Bitcoin and blockchain technologies, following the comprehensive overview of research perspectives and challenges for Bitcoin presented by Bonneau et al. in 2015 [BMC15], are necessary. In particular, the study and formalization of the Bitcoin protocol and its underlying Nakamoto consensus has seen significant advances in recent years (e.g. [KP15], [GKL16], [BPS16a]) that are not yet systematically exposed. Recent work provides a broad overview of different consensus mechanisms in the context of blockchain technologies [BSAB17], however we feel that a more in-depth analysis of the relationship between Nakamoto consensus and previous approaches to Byzantine consensus is still outstanding. We hereby narrow this gap by relating research towards Nakamoto consensus to other key insights and aspects on the topic of consensus. Consensus is a fundamental building block in fault tolerant and distributed computing, and the guarantees a consensus protocol provides can greatly impact the overall security and reliability of (distributed ledger) systems [CV17]. Bitcoin promises to solve the double spending problem in a distributed, peer-to-peer environment without the necessity to rely on a trusted third party by enabling participants to reach (eventual) agreement on the state changes to a shared transaction ledger. Nakamoto consensus hence lies at the core of this system. Without an in-depth understanding of this mechanism, entire categories of newly designed systems, as well as the applications that are built on top of them, are potentially vulnerable to attacks [NG16], [CV17]. Modifications to consensus related rules, even if they appear small or straightforward, can fundamentally impact underlying incentives and greatly affect security guarantees [ZP17]. As the ecosystem around “blockchain“ has grown into a multi-billion dollar industry, severe failures could have far-reaching consequences and a long-term negative impact on the entire field. On the other hand, fundamental insights on Nakamoto consensus have already spawned novel and hybrid consensus approaches that exhibit interesting properties and characteristics, while providing the necessary frameworks to analyze and evaluate the correctness and security of such approaches. Combining aspects of “classical“ Byzantine fault tolerant (BFT) consensus protocols with Nakamoto consensus may help to address increasing concerns regarding the scalability and performance of blockchain technologies. Motivated by this emerging new area of research, we set out to paint a coherent picture of the insights and findings that have been presented on the topic of Nakamoto consensus and the fundamental mechanisms behind Bitcoin and similar blockchain protocols.