Knowledge-based adaptive channel estimation for digital storage channels

To prevent misequalization in digital magnetic recording, we propose a method for fast channel response estimation. Utilizing the a priori knowledge about the general shape of the transition response, the channel identification problem is reduced to estimating two parameters, namely gain and the pulse width at half of the peak magnitude. This simplified estimation method has much faster convergence rate than the direct LMS approach. Once the algorithm converges, the estimates can be used to select an appropriate set of equalizer coefficients or to modify the decoder parameters