A new interpolation lattice structure

This interpolation lattice is asymmetric in the sense that the number of past and future values linearly weighted to estimate the current value is not necessarily identical. The asymmetric interpolation lattice is computationally efficient and flexible to implement. It also forms a bridge between forward prediction, backward prediction, and symmetric interpolation, which can all be viewed as special cases of asymmetric interpolation. As a result, the asymmetric interpolation lattice provides a broader interpretation and a more thorough understanding of the linear prediction and linear interpolation theories. The interpolation lattice structure may be useful in the field of data compression. Although interpolation needs more computing power than prediction does, it reduces more temporal redundancy. As a result, a higher degree of data compression can be achieved by interpolation than by prediction.<>