For a robust optimization of adaptive signal processing algorithms, errors and uncertainties of the underlying system model are described explicitly and incorporated into an optimization problem. Together with methods from estimation theory, this paradigm is applied to optimize four principal tasks of the physical layer in a wireless communication system with multiple antennas: Robust channel estimation, robust prediction of band-limited signals, estimation of spatial and temporal channel correlations, and linear/nonlinear precoding (predistortion) of the data at the transmitter for the broadcast channel. Optimal as well as suboptimal approaches of a reduced complexity are developed systematically. For precoding, e.g., Tomlinson-Harashima precoding, optimization criteria are designed which are based on imperfect channel state information or, moreover, only on the knowledge of the channel correlations.
«
For a robust optimization of adaptive signal processing algorithms, errors and uncertainties of the underlying system model are described explicitly and incorporated into an optimization problem. Together with methods from estimation theory, this paradigm is applied to optimize four principal tasks of the physical layer in a wireless communication system with multiple antennas: Robust channel estimation, robust prediction of band-limited signals, estimation of spatial and temporal channel corre...
»