A generalized data discrimination scheme using kernel machines in the presence of pattern-dependent noise
Source abstract
Abstract
Abstract This paper describes an application of the nonlinear support vector machines (SVMs) on magnetic recording channels in order to reduce performance deterioration due to pattern-dependent noise. It has been confirmed that SVMs make simple and direct modification of log-likelihood ratios possible. Consequently, at a bit error rate of 10 −4 , this method can improve a bit error rate of the pattern-dependent noise by one order in cooperation with proposed iterative decoding.
Source: openaire · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Blind Source Separation Techniques · Evolutionary Algorithms and Applications · Error Correcting Code Techniques
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (12)
Show source-provided references
1. Vapnik · 1998
2. Cristianini · 2000
3. Suykens · 2002
4. Mita · IEEE Trans. Magn. · 2004 · 40 · 21910.1109/TMAG.2003.821193 ↗
5. A. Dholakia, E. Eleftheriou, T. Mittelholzer, Iterative detection/decoding techniques for the magnetic recording channel, TMRC 2000 Digest, D-6, August 2000.
6. A.V. Kuznetsov, Coded modulation schemes with turbo and list detection for high-order PR channels, TMRC 2000 Digest, E-1, August 2000.
7. T.R. Oenning, J. Moon, A low density generation matrix interpretation of parallelly concatenated single bit parity codes, TMRC 2000 Digest, E-4, August 2000.
8. Y. Wu, J.R. Cruz, Noise predictive turbo systems, TMRC 2000 Digest, E-5, August 2000.
9. S. Mita, H. Matsui, Current and future schemes for securing data reliability in HDD, IEICE Technical Report on Magnetic Recording, July 2003.
10. Bahl · IEEE Trans. Inf. Theory · 1974 · 2010.1109/TIT.1974.1055186 ↗
11. R.J. McEliece, D.J.C. Mackay, J. Cheng, IEEE J. Sel Areas, Commun. (1998) 140–152.10.1109/49.661103 ↗
12. Mackay · IEEE Trans. Inf. Theory · 1999 · IT-45 · 39910.1109/18.748992 ↗