“…In [16], Ding et al presented a hierarchical identification algorithm to estimate the unknown parameters of the lifted state space models for general dual-rate multivariable systems. The hierarchical identification principle also has been used to identify multivariable systems described by the input-output representations, the hierarchical least squares identification methods and the hierarchical gradient-based methods for multivariable systems and their performances were discussed in [17][18][19]. Beside these contributions, there are many other effective methods developed for multivariable system identification, for example, the expectation maximization algorithm and the maximum likelihood approach [20,21], the multi-innovation based algorithms [22][23][24][25][26][27][28], the asymptotic methods [29], the iterative methods [30][31][32][33][34], the correlation technique based methods [35] and the stochastic approximation algorithm [36].…”