A New Class of Efficient Adaptive Filters for Online Nonlinear Modeling

Nonlinear models are known to provide excellent performance in real-world applications that often operate nonideal conditions. However, such require online processing be performed with limited computational resources. To address this problem, we propose a new class of efficient nonlinear for applications. The proposed algorithms based on linear-in-the-parameters (LIPs) filters using functional link expansions. In order make adaptive (FLAFs) efficient, low-complexity expansions and frequency-domain adaptation the parameters. Among family algorithms, also define partitioned-block FLAF (FD-FLAF), whose implementation is particularly suitable modeling problems. We assess compare FD-FLAFs different providing best possible tradeoff between complexity. Experimental results prove can considered as an effective solution applications, acoustic echo cancellation, even presence adverse conditions availability