Codes on graphs: normal realizations

  title={Codes on graphs: normal realizations},
  author={G. David Forney},
  journal={2000 IEEE International Symposium on Information Theory (Cat. No.00CH37060)},
  • G. Forney
  • Published 25 June 2000
  • Computer Science
  • 2000 IEEE International Symposium on Information Theory (Cat. No.00CH37060)
Wiberg et al. (see European Transactions on Telelecommunications, vol.6, p.513-25, Sept./Oct. 1995) proposed graphical code realizations using three kinds of elements: symbol variables, state variables and local constraints. We focus on normal realizations, namely Wiberg-type realizations in which all symbol variables have degree 1 and state variables have degree 2. A natural graphical model of a normal realization represents states by leaf edges, states by ordinary edges, and local constraints… 

Codes on graphs: constraint complexity of cycle-free realizations of linear codes

  • G. Forney
  • Computer Science
    IEEE Transactions on Information Theory
  • 2003
Cycle-free graphical realizations of linear codes generalize trellis realizations and can yield reductions in decoding complexity even when they do not reduce constraint complexity.

Constraint Complexity of Realizations of Linear Codes on Arbitrary Graphs

  • N. Kashyap
  • Computer Science, Mathematics
    IEEE Transactions on Information Theory
  • 2009
The vertex-cut bound, and the notion of "vc-treewidth" for a graph, which is closely related to the well-known graph-theoretic notion of treewidth, are introduced and derived to derive tight lower bounds on the kappa-complexity of any realization of C on G, which enable us to conclude that good error-correcting codes can have low- complexity realizations only on graphs with large vc-Treewidth.

The Tradeoff Between Cyclic Topology and Complexity in Graphical Models of Linear Codes

A new bound is introduced the tree-inducing cut-set bound which can be viewed as a generalization of the square-root bound to graphical models with arbitrary cyclic topologies.

Codes on Graphs : Observability , Controllability , and Local

This paper investigates properties of realizations of linear or group codes on general graphs that lead to local reducibility. Trimness and properness are dual properties of constraint codes. A

Syndrome realizations of linear codes and systems

  • G. Forney
  • Computer Science
    Proceedings 2003 IEEE Information Theory Workshop (Cat. No.03EX674)
  • 2003
It is shown that syndrome realizations are essentially the same as the restricted state realizations of Forney, G.D., Jr. (see IEEE Trans. Inf. Theory, vol.47, p.520-48, 2001), which, in some respects, appear to be more fundamental.

Codes on Graphs: Observability, Controllability, and Local Reducibility

This paper investigates properties of realizations of linear or group codes on general graphs that lead to local reducibility and finds that trimness and properness are dual properties of constraint codes that are locally reducible.

Iterative Decoding of Codes on Graphs

This report furthers the understanding of the failures in iterative decoders for the binary symmetric channel, and proposes an algorithm that aims to alleviate this drawback by constructing an equivalent graph representation that is free of four cycles.

Conditionally Cycle-Free Graphical Models for Coset Codes

Examples of a number of families of codes - including first-order Reed-Muller (RM) and the Delsarte-Goethals codes - are provided for which the proposed procedure yields optimal soft-in soft-out (SISO) decoding algorithms that are less complex than the best known trellis-based algorithms.

The Complexity Limits of Graphical Models for Linear Codes

The tree-inducing cut-set bound (TI-CSB) provides a characterization of the tradeoff between complexity and density in the family of graphical models for a given code, and can be used to show that an r th-root complexity reduction requires the introduction of at least r(r - 1)/2 cycles thus generalizing the square-root bound to graphical models with arbitrary cyclic topologies.

Convolutional Codes on Trees and Cayley Graphs

This work generalizes the notion of discrete-time index sets to in nite regular trees and introduces a generalization of conventional tail-biting to deal with the eÆcient termination of convolutional codes on trees.



Codes and Decoding on General Graphs

It is showed that many iterative decoding algorithms are special cases of two generic algorithms, the min-sum and sum-product algorithms, which also include non-iterative algorithms such as Viterbi decoding.

Factor graphs: constructions, classification, and bounds

  • R. KotterA. Vardy
  • Computer Science, Mathematics
    Proceedings. 1998 IEEE International Symposium on Information Theory (Cat. No.98CH36252)
  • 1998
This work introduces the notion of trellis formations, and obtains a classification of factor graphs into several types, based on this notion, and describes a general construction of local behaviors for trellIS formations, which generalizes the product construction for conventional trellises.

On the optimality of solutions of the max-product belief-propagation algorithm in arbitrary graphs

It is shown that the assignment based on a fixed point of max-product is a "neighborhood maximum" of the posterior probabilities: the posterior probability of the max- product assignment is guaranteed to be greater than all other assignments in a particular large region around that assignment.

Replication decoding

Two specific problems are discussed: the use of previous decisions, which leads to a weighted generalization of feedback decoding, and the extension of replication decoding to nonsystematic codes.

Coset codes-II: Binary lattices and related codes

  • G. Forney
  • Computer Science
    IEEE Trans. Inf. Theory
  • 1988
The family of Barnes-Wall lattices and their principal sublattices, which are useful in constructing coset codes, are generated by iteration of a simple construction called the squaring construction, and are represented by trellis diagrams that display their structure and interrelationships and that lead to efficient maximum-likelihood decoding algorithms.

Codes and iterative decoding on general graphs

The main thesis of the present paper is that, with respect to iterative decoding, the natural way of describing a code is by means of a Tanner graph, which may be viewed as a generalized trellis.

The dynamics of group codes: State spaces, trellis diagrams, and canonical encoders

If C is complete and strongly controllable, then a minimal encoder in controller canonical (feedbackfree) form may be constructed from certain sets of shortest possible code sequences, called granules.

Factor graphs and the sum-product algorithm

A generic message-passing algorithm, the sum-product algorithm, that operates in a factor graph, that computes-either exactly or approximately-various marginal functions derived from the global function.

Convolutional codes over groups

The first part, which covers fundamental notions such as minimality and observability, is semi-tutorial in that most definitions are already standard (within the modern behavioral theory), as are some of the formally stated results.

The generalized distributive law

Although this algorithm is guaranteed to give exact answers only in certain cases (the "junction tree" condition), unfortunately not including the cases of GTW with cycles or turbo decoding, there is much experimental evidence, and a few theorems, suggesting that it often works approximately even when it is not supposed to.