Animating Rural Broadband

Apple shocked Wall Street this week by setting new revenue records for iPhone, Services, Mac and Apple Watch sales during the 2016 holiday season and saw its stock jump 6% in a day. Facebook surprised the street with better than expected increases in EPS, revenue, and active users, getting an after-hours bump of 3%. Advances for both companies are likely to stress video, virtual reality, and augmented reality. 

Meanwhile, AT&T has announced plans to trial a Evolution 5G, a pre-5G standard system, in Austin and Indianapolis with top speeds of 400 Mbps. Standard 5G is expected to hit 1 Gbps under ideal conditions.

FCC Chairman Ajit Pai announced the formation of a Broadband Deployment Advisory Committee to speed up rural broadband upgrades, and members of Congress would like to ensure that broadband is included in any new infrastructure plan. How do these things fit together? Let’s see.

Enabling the Next Killer App

It’s no particular surprise that mobile video figures prominently in the future of communicating apps. The transition from words typed on a keyboard to video images captured by cameras has been a strong trend since the 1990s, when Mark Cuban made his first billion from the Broadcast.com IPO. We’ve come a long way from grainy video in a tiny window on a desktop PC.

Apple and Facebook are joined at the hip because most Facebook use is on smartphones. Mobile broadband networks join the party because these phones need advanced 4G and Wi-Fi networks for real-time use. 5G unlocks new capabilities because it enables bi-directional video streaming between locked-down infrastructure servers and mobile cameras.

5G needs the new spectrum identified in the Mobile Now Act: federal spectrum below 6 GHz and millimeter wave spectrum not currently in use. And it also needs the streamlined permitting and “dig once” procedure stipulated in the bill. But that’s not all.

Broadband Infrastructure Innovation

5G also needs massive investment in backhaul, the network elements that connect the mobile first mile to the rest of the Internet. Switching centers and Internet Exchanges knit privately owned networks into the amorphous mesh that is the Internet. Conventional wisdom dictates that backhaul must be optical fiber. While this is the ideal, it’s not always cost-effective or even necessary.

Just as T-Mobile built its initial network on a backhaul fabric mainly made from microwave links, much of the 5G backhaul will be wireless. If things go according to plan, some backhaul will even use a new iteration of broadband over power lines (BPL.) This system doesn’t use copper powerlines for information transmission per se. The wires serve as a sort of radio guide that steers signals with electromagnetic fields generated by power transmission.

So there are at least two ways to build a brand-new infrastructure: you can spend massively to deploy old technology such as concrete, steel, and fiber in new places. Or you can invent new technology that addresses the reasons why unserved areas have been passed by. Rural infrastructure doesn’t just suffer from high initial costs, it’s also expensive to maintain, support, and operate. All of that is important.

Stimulating Rural Broadband

There’s a growing movement toward spending lots of money on rural broadband. This is understandable given the fact that rural broadband is way behind urban broadband in performance and reliability. There’s also a mood among politicians to court rural voters for all the good reasons. And there’s the fact that rural America is where the food comes from, and in a curious way better broadband probably means better bread. Modern agriculture is a very high-tech affair, so food security depends on networking.

There are good ways and bad ways to dispense subsidies. In this case, subsidizing the development of novel technologies may get us closer to the desired state of affairs than spending more money on technologies that aren’t a great fit for the problem. If subsidies plus traditional technologies can solve the rural broadband problem today, we wouldn’t have a problem to solve.

We want to create a state of affairs in which everyone on the US has access to affordable, high-quality broadband on networks that are affordable to operate, maintain and improve. This goal implies heavy use of wireless technologies that don’t have the vulnerability to weather and natural disasters that wires do. And it also implies reconfiguration of the brick-and-mortar parts of the broadband infrastructure so that the dollars we do have to spend on wires go as far as possible.

Standard Topologies

Defining a standard implementation template for upgraded rural networks is a fun exercise. One model is a wireless first mile that includes both fixed location service for homes, offices, and farms and mobile for smartphones and mobile equipment (such as combines). This means antennas every five miles, most of which will use wireless backhaul to form a mesh.

After some number of wireless repeaters, the mesh converges on a regional network operation center. After some number of NOCs, there’s a convergence to a central NOC and then onto to an interconnection center or two for each state. Interconnection centers are probably carrier-neutral exchanges that allow connections to be made relatively close to end users.

If you’re Skyping your cousin in the next county, the session doesn’t need to leave the state. But all the carriers in the state probably can share a fiber bundle that goes to the nearest Internet Exchange. IXs are out of state in most cases as there are only a dozen major ones in the entire lower 48. So each state outside of New England needs one or two switching/co-location centers. These centers must be on the fiber highway to the Internet Exchanges.

Operator Business Models

There are lots of ways to structure a rural carrier as a business: co-ops, departments of electric utilities, small businesses, and parts of big businesses. The choices depend on incumbency in most cases, but as these networks become cheaper to maintain, economies of scale will tend to drive mergers and acquisitions. So it’s probably wise to create stand-alone business units that can be easily bought and sold.

I don’t have any special expertise regarding ideal business models, but simply wish to point out the fact that diverse structures are the present-day reality. Broadband is an expert business that can’t be done over the long haul by just anyone.

So there you have it, a sketch for better rural broadband to support emerging applications without breaking the bank. Does this work? Leave your comments and let’s hash it out.