A Finite-Length Algorithm for LDPC Codes Without Repeated Edges on the Binary Erasure Channel

Free registration required

Executive Summary

A Low-Density Parity-Check (LDPC) code is a block code defined by a sparse parity-check matrix, H. LDPC codes are typically decoded iteratively with message-passing decoding. The Tanner graph, T, of an LDPC code is a bi-partite graph which represents the parity-check matrix. This paper considers the performance, on the binary erasure channel, of Low-Density Parity-Check (LDPC) codes without repeated edges in their Tanner graphs. A modification to existing finite-length analysis algorithms is presented for these codes.

  • Format: PDF
  • Size: 90.75 KB