Absolutely not announced but I'm sure they made the go decision. Recent actions have made clear they are going forward to 30-50M homes, probably 2019-2022. This will require 100.000's of thousand of small cells and possibly over a million. They want the best network in America, probably unmatched in any other part of the world as well.
My conviction comes from my belief that the top people at Verizon are smart and knowledgeable. They've taken several actions that would be stupid if the big build wasn't decided. McAdam and team are the opposite of stupid.
Senior executives have made a dozen statements pointing the way. Verizon is a world leader in 5G research, with eleven cities mostly installed and about to go live. (Fixed only, upgrading to mobile as soon as it is ready ?2020-2021.) The technology has improved so much in the last year the engineers are confident. The cost is coming in lower than expected. Verizon can expect some crucial marketing and lobbying advantages. Everything has come together to make this the logical move for Verizon.
Verizon's 5G network - What
Homes will have an outside antenna, measured in inches, not feet. That the antenna is larger than an entire cell phone is one reason the fixed connections are easier than the mobile ones. Target speed is 5 gigabits shared, 1 gigabit to almost everybody most of the time. A wire will normally terminate in an Ethernet jack inside.
The actual signals will travel over 28 GHz, where Verizon has eight times as much spectrum as they are using today. The available spectrum allows higher speeds. They will probably start at 5 gigabits (shared) and I have seen 20 gigabit radios almost ready for market. Most customers with a good connection will see 1 gigabit almost all the time. In practice, it's very rare to see many users running 100 meg at the same time, much less 500-1000.
The radios will also support advanced LTE, with speeds that may go up to a gigabit, shared. The LTE can be put to work immediately, providing much-needed capacity. The first wave probably will be 200,00-300,000 small cells. A national network will probably require a million.
Verizon and everyone else so far is using fiber to connect to a Cloud RAN: dozens or hundreds of relatively simple cells will be connected to one big brain. NG-PON2 is the likely fiber choice. Ericsson thinks microwave backhaul has a place.
The travel back and forth to the Cloud slows things down but the savings look to be enormous. One ms looks to be a dream more than a desired product. Even connected cars and virtual reality work well at 5-10 ms I haven't spoken to a telco convinced. While 1 millisecond latency is possible and widely promoted, all the telcos who have announced are building for 5-10 milliseconds. It costs as much as double to deliver 1 ms, because you need to have a great deal of intelligence in every node.
Software Defined Networking and Network Function Virtualization will be every.