Master Thesis

Implementing asymmetric Byzantine consensus

Most protocols for Byzantine consensus used today operate under the assumption that the number of faulty processes is limited and every process in the system shares this view. This means that trust in the network is symmetric, in the sense that all processes adhere to the global assumption about the number of faulty processes.

In certain blockchain networks, however, trust is expressed in a more flexible way because they do not have global knowledge of all participants. Motivated by this situation, the Ripple and the Stellar blockchains have introduced heuristic models for flexible trust. In past research, we have formulated asymmetric Byzantine quorum systems [1] to capture such assumptions in a sound and modular way. Later we have also formulated asymmetric consensus protocols [2].

This thesis creates an implementation of an asymmetric consensus protocol. It should start from a well-established code base that implements a prototype of Byzantine consensus in the standard model, with symmetric trust.


[1] Asymmetric trust

[2] Asymmetric consensus

Contact Luca Zanolini or Orestis Alpos for more information.

Nature of the project: Theory 20%, Systems 80%.