A unified approach to the steady-state analysis of quantized adaptive filtering algorithms

The steady-state performance of adaptive filters can significantly vary when they are implemented in finite precision arithmetic, which makes it vital to analyze their performance in a quantized environment. Such analyses can become difficult for adaptive algorithms with nonlinear update equations. This paper develops a new feedback approach to the steady-state analysis of quantized adaptive algorithms that bypasses many of the difficulties encountered in traditional approaches. In so doing, we not only re-derive several earlier results in the literature, but we often do so under weaker assumptions, in a more compact way, and we also obtain new results