It is challenging to select a representative set of simulations that demonstrate the relationships between the methodologies, the configurations and the results, given the vast number of parameters. In order to concentrate on a few aspects which we considered interesting, we have mostly used small and uniform settings in the simulations presented in this section. We do not claim our network models capture all the fundamental characteristics of the Internet, but we believe that the combination of random networks and the geographic networks provides a wide enough spectrum to give us some confidence in the general utility of the methods. Throughout the section, readers are referred to Table 5.1 for the definitions of the parameters. Unless otherwise mentioned, we use the following default parameter values: the random graph scale is 100 by 100 units; the probability of network interconnection is 1.0 for both network configurations; the region size is 10 units and the co-location vicinity is two units in the random graph; the probability of including cities in the regional networks is 0.6 and 0.8 for national networks in the geographic networks.