Applause from Trevor Linney of BT. Smalls cells - WiFi or LTE - are a crucial part of the "beyond 4G" networks. It's so expensive to run fiber for backhaul no carrier in the West has a large network of freestanding small cells. (The WiFi First networks in Europe have over 10M connected. Carrier small cells are in the tens of thousands.)
LTE is starting to move to 3 carrier, 450 megabit radios, so requires plenty of speed. Over very short distances, G.fast can deliver 500 down, 200 up to an LTE cell. It's a natural way to connect a few cells in smaller office buildings.
WiFi First poses an existential threat to telcos counting on usage levels. LTE is usually charged, WiFi usually not. The profitability of telcos in the next decade will be profoundly affected by whether they keep customers away from the less profitable WiFi.
Free in France, Softbank and others have literally millions of small cells, home gateways that don't require added backhaul, Turning on a second SSID from all willing homes is by far the best, fastest and cheapest way to deliver more wireless capacity,
There are many locations where G.fast might do the job and save the cost of fiber. Mostly high rise business districts that need several small cells for some building. AT&T has been doing a fair amount of DAS for locations like that but the cost is high. There are complications to solve, especially because efficient small cell operation requires a substantial control plane. David Chambers has an interesting interview with Airspan on their LTE small cell testing with Softbank. They look forward to attaching 5 small cells and centrally manage traffic. In some cases, they can double speed that way.
There's a war on between most of the big telcos and those building capacity more efficiently using WiFi and backhaul in place. Small cells are among the tools.
Multi-gigabit access via copper
Cost-effective ultra-broadband access based on G.fast standard
23 March 2015, Heidelberg, Germany. Celtic-Plus is launching a 4.4 million euro project to explore multiple-gigabit copper access based on G.fast, a digital subscriber line (DSL) standard for the local loop. The Gigabits Over the Legacy Drop (GOLD) project will initiate the planned second version of the G.fast standard and boost its usability in dense city areas. The goal is to develop alternative backhauling options based on copper instead of fibre. This could lead to significant cost reductions in the network, particularly within dense urban areas in Europe.