A State-Space Approach to High-order Adaptive Band-Pass Filtering

This paper proposes a new method of high-order adaptive band-pass filtering for detection of narrowband signals in white noise. In our proposed method we realize a desired adaptive filter in state-space form, which leads to high-accuracy detection of narrowband signals under finite wordlength constraints. We achieve this goal by describing a gradient-based adaptive algorithm for our previously proposed tunable band-pass state-space filters. It is demonstrated that our adaptive state-space filters show the robust convergence behavior against quantization effects, whereas the convergence behavior of the conventional direct-form adaptive filters is very sensitive to the quantization effects.