Deterministic Blind Identification of IIR Systems With Output-Switching Operations

In this paper, a deterministic blind identification approach is proposed for linear output-switching systems, which are modeled by multiple infinite impulse-response (IIR) dynamic functions. By adopting a new over-sampling strategy, the concerned single-input-single-output (SISO) output-switching system is equivalently transformed into a time-invariant multi- input-multi-output (MIMO) system. Further, by exploring the mutual relations among the multiple inputs, the time-invariant MIMO system model and subsequently the output-switching system model are identified uniquely up to a scalar constant using the proposed identification approach. Sufficient identifiability conditions are provided for output-switching systems and numerical simulations are carried out to validate the proposed approach.