Method of identifying networks communities based on the Green's function of the network

Community identification has been an active area of research in complex 
networks. Many topological and dynamical properties of complex networks  are
defined by assuming that most of the transport on the network flows  along the
shortest  paths http://en.wikipedia.org/wiki/Centrality.  However there are
scenarios in which non shortest path are used to reach  network destination.
Thus the consideration of the shortest paths only  does not account for the
global communicability of complex network. New measures of the communicability
of a complex networks has proposed in 
http://www.ncbi.nlm.nih.gov/pubmed/18517465. Communicability is then view as the
Green function of the networks. The correlation  between the node degree and the
communicability (Green's function) is used in order to investigate the
structure-dynamic relationship in complex network. It has been shown in 
http://www.ncbi.nlm.nih.gov/pubmed/18517465 how communicability, or  Green
function can be used to identify network communities. The method  use the
spectral decomposition of the Green function.