Model-based interrelations of adaptive filter algorithms in acoustic echo control

The design of fast learning and yet robust adaptive filters for acoustic echo control in hands-free voice communication has attracted the research community for decades. This paper presents a fresh and unified view of adaptive filter algorithms in order to change and to generalize some of the previously well-established paradigms. The unified understanding is achieved by introducing state-space dynamical echo path models which also resemble the real world characteristics of hands-free systems.