A Deterministic Construction and Density Evolution Analysis for Generalized Product Codes
Other conference contribution, 2016

Generalized product codes (GPCs) are extensions of product codes (PCs) where code symbols are protected by two component codes but not necessarily arranged in a rectangular array. In this tutorial paper, we review a deterministic construction for GPCs that has been previously proposed by the authors together with an accompanying density evolution (DE) analysis. The DE analysis characterizes the asymptotic performance of the resulting GPCs under iterative bounded-distance decoding of the component codes over the binary erasure channel. As an application, we discuss the analysis and design of three different classes of GPCs: spatially-coupled PCs, symmetric GPCs, and GPCs based on component code mixtures.

braided codes

Deterministic generalized product codes

spatially-coupled codes

bounded-distance decoding

random graphs

staircase codes

Author

Christian Häger

Chalmers, Signals and Systems, Communication, Antennas and Optical Networks

Henry D. Pfister

Alexandre Graell i Amat

Chalmers, Signals and Systems, Communication, Antennas and Optical Networks

Fredrik Brännström

Chalmers, Signals and Systems, Communication, Antennas and Optical Networks

Erik Agrell

Chalmers, Signals and Systems, Communication, Antennas and Optical Networks

Signal Recovery: Compressed Sensing meets Coding Theory

Swedish Research Council (VR) (2011-5961), 2012-01-01 -- 2015-12-31.

Areas of Advance

Information and Communication Technology

Subject Categories

Telecommunications

Communication Systems

DOI

10.3929/ethz-a-010645440

More information

Latest update

8/29/2019