Prikkelanalyse in toestemmingsloze blockchains

Abstract

In recent years, permissionless blockchains have emerged as the foundation for a wide range of decentralized applications, including cryptocurrency management, decentralized finance, auctions, voting systems, digital ownership, and more. The reliability of these applications depends on the security guarantees offered by the underlying blockchain protocols. Although prominent permissionless blockchains often base their security on formal analyses and claims, these assessments typically rely on simplifying assumptions, such as the existence of an honest majority, or overlook critical factors, like the influence of reward distribution on system security. Permissionless blockchains incentivize their participants, referred to as miners or validators, by rewarding them with cryptocurrency to encourage active participation in the protocol. The incentive mechanism plays a crucial role in maintaining the security of the blockchain by motivating participants to follow the desired, or honest, behavior. Since validators and miners are typically rational actors seeking to maximize profit, any flaw in the incentive design may motivate them to deviate from honest behavior in pursuit of an unfair advantage. Such exploitation of the incentive mechanism can lead to a destabilizing situation in which disadvantaged participants may choose either to deviate from honest behavior or to eventually abandon the network, with both possibilities threatening the blockchain's continued progress and stability. In this thesis, we analyze the incentive mechanisms behind several prominent permissionless blockchains, including Bitcoin, Cardano, and Ethereum. Our focus is primarily on potential fairness attacks that allow adversarial miners or validators to increase their profits by capturing an unfair share of blocks. As a first step, we analyze the incentive structure in Bitcoin, the pioneering Proof-of-Work (PoW) blockchain. Specifically, we examine the temporal aspects of profitability associated with one of the most well-known fairness attacks, namely selfish mining, which aims to exclude honest blocks from the longest chain. Building on this, we assess the role of Bitcoin's difficulty adjustment mechanism in mitigating such attacks. We then explore the implications of mining pool rationality for Bitcoin's security and propose novel attack vectors that could further increase an adversary's profit in a rational setting. Additionally, we investigate how upcoming Bitcoin reward halvings and the transition toward a transaction-fee era can impact Bitcoin's security by lowering the security threshold and increasing the potential for immediate profitability. As a next step, we analyze the incentive mechanisms of Proof-of-Stake (PoS) protocols. We examine longest-chain PoS protocols, such as Cardano, and show that due to the block proposer predictability inherent in PoS systems, the selfish mining attack can be even more destructive in this context. We also analyze Ethereum as a voting-based PoS protocol. These protocols are susceptible to block reorganization attacks, in which an adversarial validator attempts to exclude blocks with weak voting support from the canonical chain. We identify a vulnerability in Ethereum's reward mechanism that results from the centralized distribution of vote rewards. If exploited, this vulnerability can lead to both short- and long-range reorganization attacks. To address this issue, we propose a fairer and more decentralized reward mechanism for Ethereum. We hope the insights learned from this thesis will help identify potential incentive-based vulnerabilities in permissionless blockchains and contribute to the design of fairer blockchain protocols.