Title :
Signal-to-noise ratio degradation with channel mismatch [in magnetic recording]
Author_Institution :
Dept. of Electr. Eng., Minnesota Univ., Minneapolis, MN, USA
fDate :
11/1/1991 12:00:00 AM
Abstract :
Sensitivities of different run-length-limited coding and equalization/detection schemes to channel response variations have been investigated. Channel variations considered are due to the head-medium velocity variations and head-medium spacing fluctuations, reflected in the height and width changes in the read pulse. Given equalizers/detectors are first tuned to a specific value of PW50, the half-height width of the read pulse. The signal-to-noise ratio degradation was observed as the actual PW50 is varied around this nominal value. The results show that the sensitivity varies widely across different detectors. For example, the partial response maximum likelihood (PRML) schemes, which achieve large performance advantages over peak detection under ideal conditions, are highly sensitive to channel variations and perform worse than peak detection as the PW50 mismatch becomes large. On the other hand, the recently introduced fixed-delay tree search with decision feedback (FDTS/DF) exhibits good resistance to channel variations while achieving favorable nominal performance
Keywords :
codes; equalisers; magnetic recording; signal detection; signal processing; PRML schemes; PW50 mismatch; channel mismatch; channel response variations; decision feedback; equalization/detection schemes; fixed-delay tree search; head-medium spacing fluctuations; head-medium velocity variations; magnetic recording; partial response maximum likelihood; run-length-limited coding; signal-to-noise ratio degradation; Degradation; Detectors; Equalizers; Feedback; Magnetic recording; Matched filters; Maximum likelihood detection; Pulse shaping methods; Signal to noise ratio; Space vector pulse width modulation;
Journal_Title :
Magnetics, IEEE Transactions on