Abstract
This brief proposes a recursive constrained sine second-order error
promoting adaptive (RCSSOEPA) algorithm. Compared with classical
recursive method, the RCSSOEPA algorithm can achieve better steady state
performance in impulsive noise. In general, the sine second-order error
(SSOE) is constructed to devise a new recursive constrained
adaptive-filtering within the least-squares framework for solving linear
constrained optimization problems. The mean-square (MS) stability of the
RCSSOEPA and its theoretical instantaneous MS deviation under Gaussian
and non Gaussian noise are analyzed, numerically investigated and
discussed in detail. Simulated results are reported to give a
comfirmation of the theoretical analysis, and show that the RCSSOEPA
outperforms recent developed constrained adaptive filtering algorithms
in the estimation misalignment and when used for system identification
under impulsive-noise.