Consensus Methods in Blockchain Systems

Autor: Julian Debus. 
Download the article as
PDF file. More information about the the Frankfurt School Blockchain Center on the Internet, on Twitter or on Facebook.

Distributed consensus systems are a novel means of establishing consensus between multiple parties concerning some piece of information, such as ownership of an asset. Consensus is recorded by cryptographically signing data, thus proving its authenticity. Blocks of signed data can be linked to previous blocks, thereby enforcing a temporal chain of blocks, giving rise to the term “blockchain”. Since the emergence of Bitcoin in 2008, various blockchain systems have been proposed with very divergent properties and areas of focus. With this paper, we analyze properties and limitations of the most important classes of blockchain systems, with a specific focus on how these systems establish consensus. Distributed consensus has become a highly dynamic field with much work being committed outside of traditional academia. Accordingly, many aspects of distributed consensus lack formal analysis and concrete criteria under which newly proposed systems can be analyzed.

We propose four criteria under which the effectiveness of distributed consensus methods can be analyzed under economic considerations. Based on these criteria, we show how the incentives of economic agents in a consensus system accomplish a stable consensus state. Additionally, we are able to use these criteria to derive challenges in proof-of-stake consensus.

Continue reading (PDF file).

Remarks

This working paper is an excerpt from a Master’s Thesis entitled “Applicability and Weaknesses of Blockchain-based Consensus Methods in Financial Markets”.

Julian Debus has been student at the Frankfurt School of Finance & Management. He has been supervised by Prof. Dr. Peter Roßbach, Professor of General Business Administration and Business Informatics.