I don't think you're going to get an answer for this question. anywhere
My hunch is that the problem is too ill-defined: (1) the response would depend upon the current & proposed implementation details, (2) such things dealing with the US government tend to be considered security risks (they can't / won't / can't reveal the details), and (3) there is no way to validate an answer.
You could try to break the problem into its components:
You could easily ask about ideas and concepts related to project estimation -- I'd use programmers.stackexchange.com -- though others may suggest elsewhere in the comments.
You can request links / resources to learn about large-scale implementations -- I'd use serverfault's meta (and let them migrate you somewhere else if need be).
You can look at the publicly available information for contracts, in the proposed stage, while in progress, and once completed to roughly estimate the time needed for an external corporation to complete a similar task.
With this information, you could estimate, how long it would take, but I return to my third point -- What if I just told you two weeks or fifty years. Sure you won't believe me. But without the underlying details/implementation you have no way to measure/validate my response.
That said, I doubt you'll be able to get more precise information than would be easily guessed at -- in other words, I bet that the best answer you'll get would be 10 years +/- 10 years, and what does that really tell you?