While iterative quantizers based on low-density generator-matrix (LDGM) codes have been shown to be able to achieve near-ideal distortion performance with comparatively moderate block length and computational complexity requirements, their analysis r
emains difficult due to the presence of decimation steps. In this paper, considering the use of LDGM-based quantizers in a class of symmetric source coding problems, with the alphabet being either binary or non-binary, it is proved rigorously that, as long as the degree distribution satisfies certain conditions that can be evaluated with density evolution (DE), the belief propagation (BP) marginals used in the decimation step have vanishing mean-square error compared to the exact marginals when the block length and iteration count goes to infinity, which potentially allows near-ideal distortion performances to be achieved. This provides a sound theoretical basis for the degree distribution optimization methods previously proposed in the literature and already found to be effective in practice.
The closure of the set of entropy functions associated with n discrete variables, Gammar*n, is a convex cone in (2n-1)- dimensional space, but its full characterization remains an open problem. In this paper, we map Gammar*n to an n-dimensional regio
n Phi*n by averaging the joint entropies with the same number of variables, and show that the simpler Phi*n can be characterized solely by the Shannon-type information inequalities
