Supervisor: Ulrich Schmid
Combinatorial Topology in Distributed Computing: Failure Models and Adversaries
Combinatorial Topology is a powerful tool, which has been used with great success in many research areas. In particular, in Distributed Computing, representing global and local configurations of a system as a discrete topological space has allowed researchers to grasp otherwise “hidden” properties. Previously unexplained impossibilities turned out to be natural consequences of certain topological properties, and thus allowed to generalize classic results such as the FLP consensus impossibility. Despite the many specific results obtained by means of Combinatorial Topology in the past, primarily in shared memory systems, its full potential in Distributed Computing remains largely unexplored.
In this thesis, we look forward to further expand the applicability of Combinatorial Topology. More specifically, going beyond shared memory systems, we will primarily study interesting distributed
computing problems under more general failure and communication models.