On Mon, Jan 20, 2014 at 4:16 PM, grarpamp <grarpamp@gmail.com> wrote:
... In US, besides the utilities themselves which do not give out detail info,
there are data providers for this, even today. they appear to charge on the order of $400/mo to $50,000/yr for various metro and long distance fiber information. they're US companies and presumably screen the hell out of you before order and scrutinize your usage after. you can call for a quote ;)
... I'm not saying he didn't compile that street level of information, only that he would have had to interact with well over a thousand entities and different data sets, even down to the individual deed, to even begin to extract that information.
huge amounts of this info used to be publicly available, much intentionally so. also, most of the plats are not so useful; you want to focus on the high capacity backbone links, transoceanic fibers, etc. these represent the vulnerable highly concentrated edges between "high degree nodes" when looking at aggregate traffic capacities. the degree of centralization / concentration is highest in communications, yet still significant in every other critical infrastructure.
... Outside of a few target areas like downtown NYC that he may have focused on for show, I doubt it was more than overlaid national powerpoints reminiscent of JYA's eyeball series.
it was much more than this. a combination of both an excessive collection he ggathered during an opportune window (pre 9/11), joined to novel application of graph theory to identify critical high degree nodes in these systems and estimate the impact of severing one or few links. (then further refined by others into a good paper i'll need to dig up, which shows that it is exceptionally difficult to win this defense against a motivated attack. e.g. almost all models for robust critical infrastructures assume random / natural failures and not targeted attacks by coordinated attackers.)