Finite-Length Extrinsic Information Transfer (EXIT) Analysis for Coded and Precoded ISI Channels

In this paper, we consider the classic problem of accessing coded and precoded intersymbol interference (ISI) channels, and particularly partial response (PR) channels, which are widely used in modeling magnetic recording systems. Previous work has either treated the binary precoder as an exclusive function of the channel characteristics, disregarding possible impact from the channel code, or, in performing the iterative analysis of this concatenated system, assumed an unbounded length. Our approach here is to extend the conventional infinite-length extrinsic information transfer (EXIT) charts to ISI systems of limited lengths, and to explicitly account for the effect of the channel nonergodicity. We found that short-length channel codes may exhibit distinctively different EXIT behaviors in different rate regions, and it is important to couple the precoder and the channel code. We show that infinite- and finite-length EXIT behaviors do not always agree, and hence the results obtained from the infinite-length analysis may not hold for practical short-length systems.

[1]  William E. Ryan,et al.  Concatenated codes for class IV partial response channels , 2001, IEEE Trans. Commun..

[2]  Jaekyun Moon,et al.  Iterative Decoding Based on Error Pattern Correction , 2008, IEEE Transactions on Magnetics.

[3]  Brendan J. Frey,et al.  Signal-space characterization of iterative decoding , 2001, IEEE Trans. Inf. Theory.

[4]  Sae-Young Chung,et al.  Analysis of sum-product decoding of low-density parity-check codes using a Gaussian approximation , 2001, IEEE Trans. Inf. Theory.

[5]  Emre Telatar,et al.  Finite-length analysis of low-density parity-check codes on the binary erasure channel , 2002, IEEE Trans. Inf. Theory.

[6]  G. Kramer,et al.  Code rate and the area under extrinsic information transfer curves , 2002, Proceedings IEEE International Symposium on Information Theory,.

[7]  Ajay Dholakia,et al.  Capacity-approaching codes: can they be applied to the magnetic recording channel? , 2004, IEEE Communications Magazine.

[8]  A. Orlitsky,et al.  Stopping sets and the girth of Tanner graphs , 2002, Proceedings IEEE International Symposium on Information Theory,.

[9]  R. Koetter,et al.  On the Effective Weights of Pseudocodewords for Codes Defined on Graphs with Cycles , 2001 .

[10]  Krishna R. Narayanan Effect of precoding on the convergence of turbo equalization for partial response channels , 2001, IEEE J. Sel. Areas Commun..

[11]  Andrew C. Singer,et al.  Turbo equalization: principles and new results , 2002, IEEE Trans. Commun..

[12]  John Cocke,et al.  Optimal decoding of linear codes for minimizing symbol error rate (Corresp.) , 1974, IEEE Trans. Inf. Theory.

[13]  Thomas J. Richardson,et al.  Error Floors of LDPC Codes , 2003 .

[14]  Stephan ten Brink,et al.  Convergence behavior of iteratively decoded parallel concatenated codes , 2001, IEEE Trans. Commun..

[15]  Tong Zhang,et al.  Quasi-Reduced-State Soft-Output Viterbi Detector for Magnetic Recording Read Channel , 2007, IEEE Transactions on Magnetics.

[16]  Inkyu Lee,et al.  The effect of a precoder on serially concatenated coding systems with an ISI channel , 2001, IEEE Trans. Commun..

[17]  Jeong Woo Lee The Study of Turbo Codes and Iterative Decoding , 2003 .

[18]  Jing Li,et al.  On the performance of high-rate TPC/SPC codes and LDPC codes over partial response channels , 2002, IEEE Trans. Commun..