This work studies the problem of static output feedback (SOF) infinite horizon robust model predictive control (RMPC) for linear uncertain systems with input constraints. In contrast with the existing RMPC design methods, this approach computes the vector of control actions directly from the observed variables while the system states are unmeasurable. For the unavailable states, an ellipsoidal set is used to obtain an estimation and applied in the RMPC optimization problem. This set is updated a
