Creating your own "mobile and internet service provider" is not an overly poor suggestion within itself, but how would you propose this actually happen? It will involve a massive investment on infrastructure - when I was looking to setup as an ISP about ten, fifteen years ago sinking £1,800,000 would of only just bought enough equipment to serve my local area. Sure exchanges have increased in capacities since then, but to remain competitive(not in terms of price wars, in terms of data capacities) then we'll need to be deploying upper end hardware. Then you have all sorts of regulations you must conform to requiring extra hardware - you need to minimally archive the metadata of your customers traffic for about five years or so in a lot of places for example. Then there's the licensing costs etc. Then you need technitions to actually admin the services, they tend to want payment. There's a massive list, and when you've worked through it you can actually start and then find real problems like the average citizen will likely not have an independant line installed so we'll still be moving our traffic through the conventional networks and accruing charges for doing so.
3G hardware is getting cheap as most companies are ditching theirs to install 4G. This could be snatched up at reduced cost, but would still represent a significant investment to cover one small country, let alone every country with citizens in, to a saturation that would actually be useful. Then there's the deployment. This need locating and providing with power. It might be possible to rent space on other's masts but can't assure this everywhere. Then ofc there's the associated regulations, licenses etc, and the maintainence of this network, ie: technitions etc.
This doesn't bypass the possibility of foriegn states or their spywares, that's mostly in the user's hardware. For some reason they thought it clever to fund the development of this so they are now reaping the rewards of their lack of thinking. They should give themselves a round of applause. The software is another issue entirely, and not really sensible to encompass until you've secured the hardware. Thanks to initiatives like CALEA then most base station hardware isn't entirely trustable either.
Even securty programs can be open sourced - if that represents some sort of threat then your security model is obvioiusly flawed and this would be unsafe to deploy. Really before worrying about software you want to create trustable hardware. Then you stand some chance of the software being able to operate securely.
In the long term scheme of things currency simply doesn't make sense - it's just a retardation imposed upon by the current environment that we only require to deal with by virtue of being in that environment. To actually have a realistic chance of providing long term mass residential habitational facilities then the only feasible method of appropriating for would involve tapping the infinite resources of the universe. When we're at this stage there's nothing that Earth has we should need. Therefore we do not need their currency.
Having a currency of our own is pretty useless, both long and short term as we'd only be able to trade between ourselves with, and really that'd make more sense to take place in the local currencies so the people involved can actually spend it. There's no viable reason why any Earth nation should accept our currency. Now or in the future. Trade to obtain things from them would naturally take place in their currencies. When we can provide for our populations habitation then we should have no need for currency by virtue of having an exponentially growing pile of resources that we really can't use fast enough - even with just giving Earth anything it wants.