Pole Attachments and Backhaul

Yesterday’s podcast is a lively and engaging discussion of 5G with Peter Rysavy, a spectrum expert whose posts have graced these pages in the past. While some hyped-up technologies seem to be slow in arriving, 5G is getter here faster than expected. While the standards aren’t expected to be done until 2020, field trials are already underway. These are parts of the standards development process because they’re providing insight about what works and what doesn’t.

While the wireless industry and most broadband users are rushing ahead to the 5G future at breakneck speed, the traditional topic of Internet access by wire isn’t dead. And it won’t be dead altogether for some time because wired networks will, along with microwave, provide essential backhaul services for 5G small cells just as they do for 4G backhaul today.

News reports indicate Google is having a hard time living up to its commitment to blanket Silicon Valley with Google Fiber because it can’t get access to utility poles. In last Friday’s San Jose Mercury News, reporter Ethan Baron charged that Google is locked into a fight with AT&T, Comcast, and PG&E (the local power company) for access to poles controlled by the three acting together as the Northern California Joint Pole Association.

Access to the Association’s poles is limited to members, and Google is not a member because it doesn’t own any poles. In some parts of the Bay Area, poles are owned by other entities such as public utilities, and Google isn’t having any problems with those poles. It also has a direct deal with AT&T and is able to use its poles as well. The Mercury News claims that Google’s issues are mainly with the fact that it’s claiming cable company benefits without actually complying with the federal Cable Act:

“While the CPUC has looked at the definition of a cable television corporation under state law, it did not consider whether Google Fiber complies with the federal Cable Act — it does not,” cable association President Carolyn McIntyre said in a March 3 email to the PUC.

“Google Fiber wants the benefits of a regulated cable corporation without the associated regulatory burdens,” she wrote.

One takeaway from this dispute is that cable companies have a right of access to utility poles that broadband-only companies do not. But the larger issue is that wired broadband is a highly capital-intensive business rather than a mom-and-pop play. If Google were to buy an inventory of poles the NCJPA would probably have little problem granting it membership. As a second article in the Mercury News explained:

Collectively, providers like AT&T and Comcast have invested hundreds of billions of dollars over the years to build out and upgrade their broadband networks, noted Bruce Leichtman, president and principal analyst at Leichtman Research Group, which focuses on the telecommunications industry. Few companies have access to that kind of money. And the more competitors they potentially face, the less likely they will be to invest that kind of money.

“It’s very hard to legislate competition and legislate spending,” Leichtman said.

One solution to the pole attachments problem is for Google to go into markets where most poles are owned by a city-owned utility or by AT&T, such as Austin. Another is the sweet deal Google got in Huntsville, where local government agreed to not just provide Google with pole attachments but to bear the costs of installing the fiber itself. FTTH fan Susan Crawford highlighted the Huntsville deal as the best model for all future networks:

The reason this is totally exciting is that this is the model that has been used with enormous success in several other cities — and could be revolutionary here in smashing the current dogma that keeps Americans overcharged and under-served.

While Americans are not really overcharged or underserved for broadband (that’s necessary political argument if you want taxpayers to subsidize your business) externalizing costs has a great deal of appeal for would-be disruptors. But it’s not really practical in most communities, as fiber installer Elliot Noss points out in Medium:

…the Golden Age is going to be far more complex and harder to achieve than [Crawford’s] article lets on: we are a long way from a truly competitive market for broadband. In fact, to get there, competition is not the most important element. For now the most important thing is to get places fibered up. How that should happen depends on local variables. But the path Google has taken in Huntsville: leaving the ownership of the broadband system in the hands of the town and renting dark fiber, rather than ceding it to the company providing the service, will happen only in the minority of places.

The Huntsville model won’t work in the normal case because most communities are not sitting on the oceans of capital it will take in most places to provide an additional gigabit wire to complement the ones already installed by the cable and phone companies. But a substantial number of communities can pay for supplemental broadband backhaul where it’s needed by mobile carriers, tower companies, and Distributed Antenna System operators.

The total system cost for a broadband network that has to connect each home with its own wire is considerably greater than that of connecting each neighborhood. A system like Verizon FiOS – where access to poles is already assured – is about $700 per home, while the cost of connecting apartments in a high rise apartment is closer to $200. The neighborhood approach would be more in line with apartment costs.

So why not make sure each neighborhood has a fiber terminal instead of using public money to connect each home? It seems to conserve dollars and make sense.