A fast affine projection algorithm for 2D filtering and linear prediction

A two-dimensional (2D) fast affine projection algorithm (FAPA) is presented for 2D adaptive filtering and linear prediction. The derivation of the proposed algorithm is based on the spatial shift invariant properties that 2D discrete time signals possess. The proposed algorithm has low computational complexity, comparable to that of the 2D LMS algorithm. The convergence speed and tracking ability of the proposed scheme are comparable to those of the higher complexity 2D RLS algorithms.