Robustness of planar random graphs to targeted attacks

In this paper, robustness of planar trivalent random graphs to targeted attacks of highest connected nodes is investigated using numerical simulations. It is shown that these graphs are relatively robust. The nonrandom node removal process of targeted attacks is also investigated as a special case of non-uniform site percolation. Critical exponents are calculated by measuring various properties of the distribution of percolation clusters. They are found to be roughly compatible with critical exponents of uniform percolation on these graphs.