Increasing densities on magnetic data storage devices leads to problems of
severe intersymbol interference (ISI), additive noise and non-linearities. Advanced
detection strategies for magnetic recording channels fall into two categories: partial
response equalization with maximum likelihood decoding and decision feedback
equalization. This study focuses on doing an adaptive all-pass forward filter for
the decision feedback channel. The decision feedback channel can be equalized by
a low-order continuous-time filter, and does not require a transversal filter with
high-precision multiplication. This results in considerable savings in both power
consumption and chip die area. One problem that has yet to be addressed is how to
adaptively set the coefficients of the all-pass filter. This thesis examines the design
and performance of an adaptive all-pass filter.
The performances in terms of the mean-squared error (MSE) of a first- and
second-order all-pass are evaluated. They are compared to a conventional FIR filter
design of various lengths. An adaptive algorithm based on the least mean-squared
(LMS) error is developed and characterized over a range of storage densities. Since
this does not require sampling of the filter input or any states of the forward filter, the system could be realized in continuous-time up to the decision device.
Numerical simulations for various data densities and noise variances are done to verify the theoretically expected performance and the adaptation behavior of the all-pass. / Graduation date: 1997
| Identifer | oai:union.ndltd.org:ORGSU/oai:ir.library.oregonstate.edu:1957/34252 |
| Date | 06 March 1997 |
| Creators | Wiedmann, Ralf |
| Contributors | Kenney, John G. |
| Source Sets | Oregon State University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis/Dissertation |
Page generated in 0.0025 seconds