Adaptive Filters
Hi-index | 0.00 |
The normalized least-mean-fourth (XE-NLMF) algorithm has a faster convergence rate and lower misalignment performance than the normalized least-mean-squares (NLMS) algorithm in sub-Gaussian noise environments. However, the XENLMF algorithm shows convergence performance degradation in highly correlated input signals. To overcome the problem, we propose an XE-NLMF algorithm with variable data-reusing. Through computer simulations, we confirmed that the proposed algorithm has a better convergence performance than the conventional XE-NLMF algorithm for colored input signals.