Sparse Signal Approximation Algorithms in a Celp Coder
This paper presents a series of sparse signal modeling algorithms implemented in a typical CELP coder in order to compare their performances at a reasonable computational load. New algorithms are proposed, based on cyclic and parallel use of a fast implementation of the optimized orthogonal matching pursuit algorithm, i.e., the recursive modified Gram - Schmidt algorithm. These algorithms yield a statistically significant reduction of signal approximation error at a controllable computational complexity. The idea of CELP coder emerged in the early 80's when the possibilities of real time implementation were limited.