Fast and Adaptive BFT State Machine Replication

Recently, Byzantine fault-tolerant (BFT) state machine replication (SMR) experiences renewed research interest with the rise of novel BFT SMR-based distributed ledger technologies (DLTs). In DLTs, BFT SMR is used as a core primitive for maintaining a consistent append-only log by committing blocks of transactions in all replicas. In comparison with Proof-of-Work, BFT can typically improve performance and energy efficiency and comes with proven system properties. BFT protocols can be used to create resilient infrastructures, but still open research questions exist and deserve further investigation. For instance, one major ongoing challenge is currently to make "traditional" BFT applicable to practical, large-scale environments. Another challenge is to assert that the BFT system suffices other requirements necessary for a practical deployment, e.g., being resilient to performance degradation or liveness-endangering attacks. The high level goal of this PhD project consists in analyzing how BFT can be optimized by making it adaptive towards its environment such as system size or geographic dispersion of replicas while hardening the resilience of the system.