AT&T has released one tiny detail which we all could and should have already inferred about its Project AirGig – that it is in advanced discussions with multiple power companies to install it and trials will be held this year – certainly in one location and possible in more.
Just to remind you, AirGig is one method for using wireless millimeter wave point to point, to power fixed line broadband connections – both commercial broadband at 1 Gbps, and backhaul for AT&T’s own small cells ramp up.
It relies on a plastic antenna which can be installed on power lines, where a more or less, unobstructed line of sight can be guaranteed, without trees or buildings getting in the way of millimeter wave signals. Communications infrastructure has always followed previous generation infrastructure, and a map of major telecommunications routes often runs along major trunk roads, railway lines (often the best place to lay a fiber), which before them paralleled canal pathways – so this is a similar transition idea. Power has to get to every house, so the poles that carry it must get within 100 feet of most homes – which is about the distance millimeter wave spectrum can travel safely. And this infrastructure is already in place, so will require only minimal installation costs, and will also reach inside cities as well as into rural communities.
Faultline Online Reporter has consistently raised the question about just how backhaul will reach the 100,000s of small cells which are envisaged in the major structural adjustments needed for 5G. Roughly for every 1 Macrocell in previous generations of cellular required, 5G will need 100 small cells, in order to gain the full benefits of the air interface design. To get this in detail try drawing a Macrocell with a diameter of 6 kilometers, and then draw a scale circular cell of 100 to 200 meters. Now see how many of these it takes to fill the same Macrocell space.
Connecting all of these need radical solutions and increasingly the one that has stood out as an economic idea is to allow all of a single country’s MNOs to use the same shared infrastructure – this is seen as the only way to afford 5G at a modest clip, rather than stretching out its implementation over 15 years. But it has not had many fans as an idea within existing MNOs.
There are more and more reports of City councils taking on the cost of building out base stations and then renting the same back to multiple MNOs on a guaranteed SLA that allows it to finance the deal – keeping debt off the books of MNOs. AT&T must now pioneer, for the entire industry, the type of contract that allows it to use Power Utility owned poles.
It can go two routes and we think we know which of them it will choose. The first is to build a robust network and sell service to ALL MNOs equally, and in this situation, an implementer like Ericsson might be responsible for the installation and for setting up finance.
Or it can build them for one MNO, on the understanding that if that MNO wants to share them, it will be the MNO’s decision. This may fetch the Utility less money, which is why AT&T is touting the idea that it has tons of Intellectual Property in this approach, suggesting that other MNOs wouldn’t be allowed to use its ideas anyway. AT&T will want to have exclusivity on small cells supported by such an approach, powerline companies will want to sell the installations over and over again, on a usage based fee. But if AT&T gains a huge timing advantage, it may be prepared to pay more handsomely for use of the poles, if they are provided on an exclusive basis.
We don’t think it would be practical for AT&T or anyone else to actually install the antennas on the powerlines. Utilities have the workforce in place trained in dealing with mid-voltage powerlines. This is a way of offloading some of that workforce’s cost onto a new revenue stream. In that sense, it is a win-win.
The IPR AT&T says it owns is in creating durable plastic antennas which don’t have to be precisely aligned for line of sight communication, but instead just have to be roughly in line with each other to manage the 10 Gbps backhaul hops along a power line. It calls these Radio Distributed Antenna Systems.
“We are looking forward to begin testing the possibilities of AT&T Labs’ invention for customers and utility companies,” said Andre Fuetsch, president of AT&T Labs and chief technology officer. “AT&T is focused on delivering a gigabit-per-second speed everywhere we can with our wired and wireless technologies.
AT&T thinks of this entire build out as 5G, although it will go into business initially without roaming or mobility, offering a fixed service first, which augments its wireline services, rather than wireless, with each hop from the powerline utility coming off-ramp, capable of scaling to 1 Gbps. This is just one way that AT&T is planning to leap into really high speed broadband, its others being G.fast and fiber.
The great advantage that Project AirGig has is that it requires no separate power connection, instead using induced power from the powerlines, which again leads us to believe that the Utility owner must bill for the service on a usage basis, not charging for power as such, but charging for a complete service, easily measured in bandwidth delivery, which uses up an impossible to measure amount of power. AT&T confirmed in this week’s statement that the system uses inductive power devices, as we had theorized a few months back.
In return for all this, AT&T has promised it will work on smart grid technologies, so that it can carry meter data and smart meter control information along the same network, killing two birds with one stone and providing a network for IoT smart energy systems.
There is one caveat, AT&T talks about the trial evaluating how such a system works in the rain, snow and high winds and work out the cost of maintenance
AT&T added that it is already marketing a 1 gigabit connection over fiber which already reaches 4 million locations in 46 markets (although mostly priced for enterprises).
In the release, AT&T talked about installing C-RAN and decided to change this from the accepted term Cloud-RAN to Centralized-RAN, which is on trial in San Francisco, where it is using streetlamps, among other things, to carry DAS and AT&T said this week it has acquired FiberTower out of bankruptcy for an undisclosed amount, along with its mmWave spectrum rights. FiberTower has assets in 24GHz and 39GHz and provides wireless backhaul. It also said that its G.fast based services would be launched in mid-2017.
Meanwhile Google said this week it would add six fixed wireless markets to its Google Fiber coverage map, after acquiring fixed wireless provider Webpass last year. Webpass has a few thousand clients in a handful of US markets in using fixed wireless service on licensed and unlicensed band, from 6GHz all the way up to 80 GHz bands.