A Signal Processing Algorithm Based on Parametric Dynamic Programming

摘要

A new algorithm for low-level signal processing is proposed based on dynamic programming principle. It is shown that it is possible to extend the dynamic programming procedure to the case of continuous variables by introducing the parametric family of Bellman functions, represented as a minimum of a set of quadratic functions. The procedure can take into account a wide range of prior assumptions about the sought-for result, and leads to the effective algorithms of data analysis.