Constraint Complexity of Realizations of Linear Codes on Arbitrary Graphs

A graphical realization of a linear code C consists of an assignment of the coordinates of C to the vertices of a graph, along with a specification of linear state spaces and linear local constraint'' codes to be associated with the edges and vertices, respectively, of the graph. The $\k$-complexity of a graphical realization is defined to be the largest dimension of any of its local constraint codes. $\k$-complexity is a reasonable measure of the computational complexity of a sum-product decoding algorithm specified by a graphical realization. The main focus of this paper is on the following problem: given a linear code C and a graph G, how small can the $\k$-complexity of a realization of C on G be? As useful tools for attacking this problem, we introduce 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. Using these tools, we derive tight lower bounds on the $\k$-complexity of any realization of C on G. Our bounds enable us to conclude that good error-correcting codes can have low-complexity realizations only on graphs with large vc-treewidth. Along the way, we also prove the interesting result that the ratio of the $\k$-complexity of the best conventional trellis realization of a length-n code C to the $\k$-complexity of the best cycle-free realization of C grows at most logarithmically with codelength n. Such a logarithmic growth rate is, in fact, achievable.

💡 Research Summary

**
This paper investigates the fundamental limits on the “constraint complexity” (denoted κ‑complexity) of graphical realizations of linear codes. A graphical realization assigns each coordinate of a linear code C to a vertex of a connected graph G, equips each edge with a linear state space, and attaches a linear local constraint code to each vertex. The κ‑complexity of a realization is defined as the maximum dimension among all local constraint codes; it directly reflects the computational load of the sum‑product message‑passing decoder associated with the realization.

The central problem addressed is: given a code C and a graph G, how small can the κ‑complexity of a realization of C on G be? While earlier work focused on tree (cycle‑free) realizations—where a unique minimal tree realization exists—and on tail‑biting trellises (single‑cycle graphs), there was no general theory for arbitrary graphs. The authors fill this gap by introducing two new analytical tools:

1. Vertex‑Cut Bound – an analogue of the classic Edge‑Cut Bound, but tailored to constraint complexity. If a set of vertices W separates G into components, the sum of the dimensions of the local constraints in each component cannot be smaller than the κ‑complexity of the whole realization. This bound provides a clean, graph‑theoretic lower limit on κ‑complexity.

2. vc‑treewidth – a graph invariant derived from a data structure called a vertex‑cut tree. A vertex‑cut tree records all possible vertex cuts of G in a tree‑like hierarchy; each node is labeled with a “vc‑width”, the total dimension contributed by the cut it represents. The vc‑treewidth of G is the minimum possible vc‑width over all such trees. This quantity is closely related to the classical treewidth, but is specifically designed to capture the behavior of constraint codes rather than merely graph connectivity.

Using these tools, the authors prove a fundamental inequality: \