Basically what a cell signal already does when a user moves between towers, but Over The Internet. So probably already patented by 30 different companies.
This is why I would love to see a cheap data only service. There is no reason for smartphones to have dedicated voice/text service when it could all be taken care of by a data connection and a VoIP provider.
This is why cheap data service can't exist right now.
Someone has to build the towers, string the cables, install the radios and antennas. Someone has to change the oil in the genset, and (depending) rotate out the diesel. Someone has to maintain the aircraft warning lamp. Someone has to handle ESD (lightning) damage. And still, someone has to deal with farmers and their backhoes.
And mitigate interference and overlap issues. And deal with routing issues. And. And. And.
There's no way for a data service to be cheap: With modern codecs, voice (which is much, much better than my first digital/non-AMPS cell phone) already uses very little data, and Youtube uses lots. Which is why unlimited voice/text service is cheap, and genuinely unlimited data is like a hen's tooth.
I used to get consistently better bandwidth with my OG Droid on genuinely-unlimited 3G than on any public hotspot, so when I was stuck in one place for awhile (selling/"donating" blood plasma, for instance) I'd just stream Netflix over 3G instead of using Biolife's carrier-grade TDM-sourced Wifi.
But in doing so I knew that I was squandering a limited resource: Actual bandwidth, aka spectral capacity.
And the only way to increase that total available bandwidth is to have more towers with smaller footprints AND maybe an institution of like-minded people who securely open up their home routers for the world to use (didn't Vodaphone do this on the other side of the pond?).
But the first case involves lots of money (see above), and the second case involves cooperation and trust and hardware that is smart enough to configure itself to deal with interference mitigation autonomously.
Both concepts can have traction and will work with existing technology, though the latter will fall apart with non-stationary users since there's a -lot- to be desired in a given Wifi client device's ability to handle roaming between multiple disparate networks.