Attack Vectors in Proof of Stake Blockchains

Abstract

Proof of Stake blockchains present a method for managing a public ledger and validating ledger data without the costly mining process associated with Proof of Work. This paper examines attack vectors in proof of stake algorithms, with particular focus towards the Ouroboros blockchain protocol. Ouroboros is one of few Proof of Stake blockchains which makes specific claims about its security, claims to have undergone academic peer review, and describes its protocol and security properties in a more rigorous white paper. This paper aims to expand upon existing research on security in Proof of Stake. This paper particularly examines Nothing-at-Stake attacks and Selfish-Mining attacks similar those proposed in the Bitcoin Blockchain, aiming to adapt the concept to Proof of Stake. The vast majority of current cryptocurrencies use Proof of Work consensus algorithm, as do the largest such as Bitcoin and Ethereum. Some attention is now being given to Proof of Stake consensus algorithms, motivated primarily by the growing electricity and capital costs associated with Proof of Work. Proof of Stake, however, is not yet accepted as a reliable or secure blockchain protocol, as insufficient research and real-world testing has been done on its security. It is the intention of this paper to further the research on Proof of Stake security, and motivate future work on the topic.