A LATTICE PREDICTOR BASED ADAPTIVE VOLTERRA FILTER AND A SYNCHRONIZED LEARNING ALGORITHM (ThuPmOR1)

Author(s) : Kenji Nakayama (Kanazawa Univ., Japan) Akihiro Hirano (KanazawaUniv., Japan) Hiroaki Kashimoto (Kanazawa Univ., Japan)

Abstract : This paper proposes a lattice predictor based adaptive Volterra filter and a synchronized learning algorithm. In the adaptive Volterra filter (AVF), the eigenvalue spread of a correlation matrix is extremely amplified, and its convergence is very slow for gradient methods. A lattice predictor is employed for whitening the input signal. Its convergence property is analyzed. Convergence is highly dependent on a time constant parameter used in updating the reflection coefficients. Furthermore, a synchronized learning algorithm is proposed, by which fast convergence and a small residual error can be achieved. Computer simulations, using colored signals and real speech signals, demonstrate that the proposed method is 10 times as fast as the DCT based AVF.

Menu