Title :
Dropout Compensation by Equalizer Selection and Kalman Filter-Based Timing Recovery
Author :
Xie, Jin ; Kumar, B. V K Vijaya
Author_Institution :
Data Storage Syst. Center, Carnegie Mellon Univ., Pittsburgh, PA
fDate :
5/1/2007 12:00:00 AM
Abstract :
When occasional dropouts occur in magnetic tape recording systems because of defects and debris, the readback signal exhibits amplitude loss, pulse widening, and peak shifts, and the corresponding entire data block may be lost. In this paper, we develop a dropout compensation method that works in conjunction with the interpolated timing recovery method. In this method, we estimate the readback signal envelope and use it as a measure of the depth of a dropout. On the basis of the estimated envelope, an equalizer can be selected from a pretrained filter bank to transform the distorted transition response in a dropout to the desired partial response target. In addition, a Kalman filter-based scheme can be used to determine the phase and the frequency update gains in the timing recovery loop. We show that, by using Kalman filter-based timing recovery, more dropouts can be compensated
Keywords :
Kalman filters; equalisers; magnetic recording; tape recorders; Kalman filter-based timing recovery; amplitude loss; distorted transition response; dropout compensation; equalizer selection; magnetic tape recording systems; peak shifts; pulse widening; readback signal; readback signal envelope; Bit error rate; Data engineering; Data storage systems; Equalizers; Filter bank; Frequency; Kalman filters; Magnetic recording; Magnetic separation; Timing; Dropout compensation; Kalman filter; equalizer; timing recovery;
Journal_Title :
Magnetics, IEEE Transactions on
DOI :
10.1109/TMAG.2007.890679