Master Thesis
Asymmetric Trust Compilers
Distributed systems are prone to failures. It is therefore necessary to make assumptions about the possible failures that might occur to be able to solve problems in these environments. In most systems, all participants must have the same assumptions, also referred to as symmetric systems. Asymmetric trust[3, 4] is a novel idea that allows participants in distributed systems to use their own failure assumptions, independent from the choices made by other processes. This more accurately models the real world, where different people have different trust notions.
Using asymmetric assumptions it is possible to express many more failure scenarios than with symmetric ones [3]. However, it has recently been shown [1, 2] that the actual scenarios where it is possible to solve problems (like reliable broadcast and consensus) are the same in both. That means that even though with asymmetric trust one can express many more scenarios, there exist no algorithms that function correctly in all the new cases. This shows an equivalence between both models. These results have been shown for crash and Byzantine failures in the scenario with a quorum setup phase, which means that the assumptions made by each process are known by all processes in the system.
In this thesis we are interested in exploring the same question in the setting without a quorum setup phase and with crash faults. In this scenario, each process only knows its own trust assumptions. We will determine if there exists some equivalency between symmetric and asymmetric trust. Possible tasks to do include developing distributed algorithms, proving their correctness, and developing impossibility proofs for distributed systems. If the thesis has a positive outcome there exists the possibility to publish the results in scientific conferences. An exposure to distributed algorithms is beneficial for this thesis but not needed. The student should have done well in most theory-based lectures offered by the university.
References
[1] Asymmetric Failure Assumptions for Reliable Distributed Systems
[2] Weaker Assumptions for Asymmetric Trust
[3] Secure Protocols with Asymmetric Trust
[4] Asymmetric Distributed Trust