Patent · US Active

Message-passing and forced convergence decoding method

US8468436B2 · kind B2 · utility

0Cited by
1References
13Claims
0Family size

Assignee

Inventor

Key dates

Filing dateFeb 11, 2008
Grant dateJun 18, 2013
Priority date
Expiry dateFeb 18, 2032

Classification

  • Technology area (CPC H)Electricity
  • CPC primaryH03M13/658
  • WIPO fieldBasic communication processes
  • WIPO sectorElectrical engineering

Abstract

The invention relates to an iterative decoding method of the message-passing type for decoding an error correcting code susceptible of representation by a bipartite graph including a plurality of variable nodes and a plurality of control nodes, said messages being expressed in terms of a log likelihood ratio. At each iteration of a plurality of decoding iterations, for each pair consisting of a variable node and a control node, a change is detected in the sign of the extrinsic information intended to be transmitted (αmntemp) as a message by said variable node to said control node relative to that transmitted (αmn) at the previous iteration, and in the event of a sign change, said extrinsic information is subjected to an amplitude reduction operation (Fred) before it is transmitted to the control node.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.