An optimized affine projection algorithm for acoustic echo cancellation

The overall performance of the affine projection algorithm (APA) is highly influenced by two important parameters, i.e., the step-size and regularization terms. In this paper, we follow a joint-optimization problem on both these parameters, in the context of a state variable model, which fits very well for acoustic echo cancellation (AEC). Thus, the proposed optimized APA achieves a proper compromise between the performance criteria, i.e., fast convergence/tracking and low misadjustment. Simulations performed in the context of AEC indicate that the proposed algorithm could be an attractive choice for this important application.