Abstract
This paper is concerned with finite impulse response (FIR) filtering with noisy input and output measurements. A new least-squares (LS) based algorithm is proposed to estimate the FIR filter coefficients. It is shown that the noise-induced bias can be removed once the variances of the input noise and output noise are obtained. A simple procedure is presented for estimating these variances by taking advantage of the FIR filter structure. The proposed LS based algorithm is easy to implement. Numerical results that illustrate the attractive properties of the new FIR filtering algorithm are presented.
| Original language | English |
|---|---|
| Pages (from-to) | IV444-IV447 |
| Journal | Proceedings - IEEE International Symposium on Circuits and Systems |
| Volume | 4 |
| Publication status | Published - 2003 |
| Event | Proceedings of the 2003 IEEE International Symposium on Circuits and Systems - Bangkok, Thailand Duration: 25 May 2003 → 28 May 2003 |