About cookies on this site Our websites require some cookies to function properly (required). In addition, other cookies may be used with your consent to analyze site usage, improve the user experience and for advertising. For more information, please review your options. By visiting our website, you agree to our processing of information as described in IBM’sprivacy statement. To provide a smooth navigation, your cookie preferences will be shared across the IBM web domains listed here.
Conference paper
Coding for higher-order partial-response channels
Abstract
Partial-response maximum-likelihood (PRML) methods are now being adopted in many digital magnetic recording systems. It is expected that as linear densities continue to increase, there will be a need to use 'extended' PRML techniques. In fact, commercial systems incorporating extended partial-response target channels, denoted EPRML and EEPRML, employing the EPR4 transfer polynomial h(D) equals 1 plus D minus D 2 minus D 3 and the EEPR4 transfer polynomial h(D) equals 1 plus 2D minus 2D 3 minus D 4, respectively, have recently appeared. Among these systems, several apply the rate 2/3, (d,k) equals (1,7) runlength-limited code, originally designed for use with peak-detection, in combination with a detector trellis structure reflecting the d equals 1 constraint. In the EEPR4 case, the d equals 1 constraint is known to provide a coding gain of 2.2 dB, unnormalized for the rate loss, relative to the uncoded channel. In this paper, we describe a nested family of code constraints, properly containing the d equals 1 constraint, intended for use on the EEPR4 channel. These constraints are shown to have the same distance-enhancing properties as the d equals 1 constraint. They permit the design of practical codes for EEPR4 that offer the same coding gain as the (1,7)-coded system, but with higher achievable code rates. The paper concludes with the construction for such a code which, having rate 4/5, offers a 20% increase over the 1,7) code.
Related
Conference paper
Unassisted true analog neural network training chip
Conference paper