Distributed estimation over an adaptive diffusion network based on the family of affine projection algorithms

In this paper, we present the distributed selective partial update affine projection algorithm (dSPU-APA), the distributed dynamic selection affine projection algorithm (dDS-APA), and dSPU-DS-APA to solve distributed estimation problem over adaptive diffusion networks. In dSPU-APA, the filter coefficients are partially updated during the adaptation. For dDS-APA, the input regressors are dynamically selected at each iteration. Diffusion cooperation schemes have been shown to provide good performance in these algorithms. The good performance of these algorithms is demonstrated by presenting several simulation results.