Hierarchical Extended Stochastic Gradient Algorithms for Single-input Mutiple-output Systems
Xiao Yong-song · Science Technology and Engineering · 2009
According to the negative gradient search principle,a hierarchical extended stochastic gradient (HESG) algorithm is derived for single-input multioutput systems with the moving average noise. In order to improve the convergence rate,a HESG algorithm with a forgetting factor is obtained by introducing a forgetting factor in the HESG algorithm.The simulation results show that the proposed algorithms work quite well.