An input-output representation is obtained for the periodically non-uniformly sampled systems by using the
discretization technique of the state space models. According to the hierarchical identification principle, the obtained
identification model is decomposed into two fictitious subsystems, one containing the parameter vector and the other
containing the parameter matrix. Taking into account the causality constraints, the subsystem with the parameter matrix
is also decomposed into several sub-subsystems, and the hierarchical least squares identification method is developed for the
non-uniformly sampled systems. The simulation results show the effectiveness of the proposed algorithm.