In this paper, a novel approach to statistical modelingof power-line channels is illustrated and its application tolow-voltage indoor power networks is analyzed in the bandwidth1–30 MHz. The proposed approach is based on the well-knownbifilar model and on a generalization of the so-called -branchnetwork topology. It is shown that our model can be exploitedto devise an efficient channel simulator predicting the meanimpedance matrix and transfer function between an arbitrarycouple of plugs in a class of indoor networks sharing multipleparameters (e.g., number of branches, minimum and maximumcable lengths, power-loading conditions). Finally, it is shown thatMonte Carlo results generated by our simulator are in good agreementwith a set of experimental data acquired in a measurementcampaign.