earticle

영어

In the orthogonal direction (OD) algorithm, the iteration direction is inconsistent with the direction that leads to the iteration error, thereby, degrading the convergence performance of adaptive filtering. In this paper, we propose a new self-adaptive orthogonal direction algorithm (called OD-DE) by introducing iteration direction error. In the OD-DE algorithm, the iteration direction is changed to be consistent with the direction that causes the iteration error, both equal to the direction vector established for input signal. Besides, we also estimate the optimal step-size for the OD-DE algorithm, to further improve the convergence performance of adaptive filtering. Last, the simulation experiments demonstrate the effectiveness of our proposed algorithm.