Optimized computational Afin image algorithm using combination of update coefficients and wavelet packet conversion
Subject Areas : journal of Artificial Intelligence in Electrical Engineeringjhila mohammadiyan 1 , Jafae Rasi 2
1 - Department of Electrical Engineering. Ahar Branch, Islamic Azad University, Ahar, Iran
2 - Department of Electrical Engineering. Ahar Branch, Islamic Azad University, Ahar, Iran
Keywords: Update Selection Coefficients, Data Selector, Set-Membership Filtering, Packet Wavelet,
Abstract :
Updating Optimal Coefficients and Selected Observations Affine Projection is an effective way to reduce the computational and power consumption of this algorithm in the application of adaptive filters. On the other hand, the calculation of this algorithm can be reduced by using subbands and applying the concept of filtering the Set-Membership in each subband. Considering these concepts, the first step is to apply an adaptive algorithm in the packet wavelet transform domain in this Article, and then using multiple decimate and independent filters instead of an adaptive filter in the corresponding packages of this structure is suggested. Then the EFI image algorithm is applied to the corresponding packages of this structure, and the three methods of data selector based on the Set-Membership filtering concept, the method for selecting the optimal coefficients for the update, and the observation selector method in the same time in all corresponding packets in the AFin algorithm apply. The simulation results show the efficiency of the proposed algorithm in terms of the convergence rate and the reduction of computational complexity.