Probabilistic processing offers a promising alternative to costly hardware-based fault mitigation, particularly in ML. However, for satellite systems affected by space radiation, the existing approaches are expected to be ineffective. This Master’s project analyzes the satellite fault scenario to provide an analytical basis for this expectation, conducts simulations of state-of-the-art probabilistic methods to identify their limitations, and derives transfers these insights towards the development and expectations of binary domain optimization, a new probabilistic processing approach currently under development.