We address the early design of complex, largescale systems by viewing them as random networks and optimizing structure over their generative parameters. In this approach, we do not seek speciﬁc topologies, but rather classes of nearoptimal networks which correspond to desirable statistical behavior, while also allowing ﬂexibility to accommodate unmodeled constraints.
Functionally, we perform the optimization of generative parameters on small networks (e.g., one hundred nodes) and use the results to design large networks (e.g., one thousand or more nodes). This approach is a computationally feasible forward design path for largescale systems.
A numerical example is given in which a network’s degree distribution is optimized for combined robustness and cost in a cascading failure scenario; the work has direct application to distributed communication systems.
For more information:
 Hummel, R., J. Taylor, and F. Hover, "Numerical Optimization of Generative Network Parameters," ASME International Mechanical Engineering Congress and Exposition (IMECE), Vancouver, November 2010. (Full Text, PDF)

The networks shown here have the same number of connections, yet the optimized network is much more robust to failure than the random network. 