Two-dimensional lattice linear prediction parameter estimation method and fast algorithm

This paper presents an extension of the one-dimensional (1-D) lattice (refl ection coefficient) technique of linear prediction parameter estimation, fi rst popularized by Burg, to the tno-dimensional (2-D) case. The resulting f ast recursive 2-D algorithm is a significant computational simplification o ver and an estimation improvement on previous attempts to extend the 1-D Bu rg linear prediction algorithm to to by exploiting some newly discovered ma trix structures. The technique presented here is useful for high resolution 2-D spectral analysis applications and the creation of high-resolution spo tlight-mode synthetic aperture radar (SI-IR) imagery, as will be illustrate d.