Adaptive combination of step sizes for distributed incremental affine projection algorithm

The distributed incremental affine projection algorithm (DIAPA) with constant step size inevitably encounters the conflict between the fast convergence rate and low steady-state misalignment. Thus, this paper proposes an adaptive combination of step sizes for distributed incremental affine projection algorithm (ACSS-DIAPA), which is updated by effectively employing a mixing parameter to combine the large step size and small step size. The mixing parameter is calculated by minimizing the mean-square deviation (MSD). The advantages of the proposed algorithm are illustrated by simulations.