The Reality of Rural Broadband

I recently saw the results of several rural surveys that probably tell the best story about the state of rural broadband. The two areas being studied are far apart geographically, but they are similar in many ways. The areas are both rural and are not near to a metropolitan area. The areas have some modest manufacturing and some modest amount of tourism, but neither in a big way. Both areas included some small towns, and a few of these towns have cable TV. And in both places, the customers in the rural area have poor broadband choices. These are not small isolated pockets of people, and the two surveys cover nearly 20,000 homes.

If you listen to FCC rhetoric it’s easy to think that rural broadband is improving – but in areas like these you can’t see it. These areas have both were supposed to get some upgrades from CAF II – but from what the locals tell me there have been zero improvements so far. The CAF program still has a few years to go, so perhaps there will be some modest improvement in rural DSL.

For now, the broadband situation in these areas is miserable. There are homes with DSL with speeds of a few Mbps at best, with some of the worst speeds hovering at dial-up speeds. One respondent to a survey reported that it took 8 hours to download a copy of Microsoft Office online.

The other broadband choices are also meager. Some people use satellite broadband but complain about the latency and about the small data caps. These areas both have a smattering of fixed wireless broadband – but this is not the modern fixed wireless you see today in the open plains states that delivers 25 Mbps or faster broadband. Both of the areas in the surveys are heavily wooded with hilly terrain, and fixed wireless customers report seeing speeds of 1-2 Mbps. There are a number of homes using their cell phones in lieu of home broadband – an expensive alternative if there are school kids or if any video is watched. There were customers who reported using public hotspots in nearby small towns. And there were a number of households, included many with school kids who have given up and who have no broadband – because nothing they’ve tried has worked.

As would be expected in rural areas, slow speeds are not the only problem. Even homes that report data speeds that should support streaming video complain that streaming doesn’t work. This indicates networks with problems and it’s likely the networks have high latency, are full of jitter, or are over-subscribed and have a lot of packet loss. People don’t really judge the quality of their broadband connection by the speed they get on a speed test, but instead by the ability to do normally expected activities on the Internet.

Many of these homes can’t do things that the rest of us take for granted. Many report the inability to stream video – even a single stream. This is perhaps the biggest fallacy in the way the FCC measures broadband, because they expect that a house getting a speed like 5 Mbps ought to be able to do most needed tasks. In real life the quality of many rural connections are so poor that they won’t stream video. Many people in these areas also complained that their Internet often froze and they had to constantly reboot – something that can kill large downloads or kill online sessions for school or work.

One of the biggest complaints in these areas was that their network only supported one device at a time, meaning that members of the family have to take turns using the Internet. I picture a family with a few school kids and can see how miserable that must be.

The surveys produced a long list of other ways that poor broadband was hurting households. Number one was the inability of people to work at home. Many people said they could work at home more often if they had broadband. A few respondents want to start home businesses but are unable to because of the poor broadband. Another common complaint was the inability for kids to do schoolwork, or for adults to pursue college degrees on line.

The problems many people reported were even more fundamental than these issues. For instance, there were households saying that they could not maintain a good enough connection to bank online or pay their bills online. There were respondents who say they can’t shop online. Many households complained that they couldn’t offload cellular data at home to WiFi, driving up their cellular bills. A number of homes would like to cut the cord to save money but can’t stream Netflix as an alternative to cable.

When you look the raw data behind these kinds of surveys you quickly see the real issues with lack of broadband. In today’s society, not having home broadband literally takes a home out of the mainstream of society. It’s one thing to look at the national statistics and be told that the number of homes without broadband is shrinking. But it’s an entirely different story when you see what that means for the millions of homes that still don’t have adequate broadband. My guess is that some of the areas covered by these surveys show as underserved on the FCC maps – when in fact, their broadband is so poor that they are clearly unserved, ignored and forgotten.

Issues Facing Cellular Networks

Cell-TowerMost networks today are under stress due to growing broadband traffic. The networks that are easily the most stressed are cellular networks and I think that there can be lessons learned in looking how mobile providers are struggling to keep up with demand. Consider the following current issues faced by cellular network owners:

Traffic Volume Growth. Around the world cellular networks are seeing between 60% to 120% annual growth in data volumes. The problem with that kind of growth is that as soon as any upgrade is made to a part of the network it is consumed by the growth. This kind of growth means constant choke points in the network and problems encountered by customers.

The large cellular companies like Verizon and AT&T are handling this with big annual capital budgets for network improvements. But they will be the first to tell you that even with those expenditures they are only putting band-aids on the problem and are not able to get ahead of the demand curve.

WiFi Offload Not Effective. For years cellular networks have talked about offloading data to WiFi. But the industry estimates are that only between 5% and 15% of data through cellphones is being handled by WiFi. This figure does not include usage in homes and offices where the phone user elects to use their own local network, but rather is the traffic that is offloaded when users are outside of their base environment. Finding ways to increasing WiFi offload would lower the pressure on mobile networks.

Traffic has Moved Indoors. An astounding 75% of mobile network traffic originates from inside buildings. Historically mobile traffic came predominantly from automobiles and people outside, but the move indoors looks like a permanent new phenomenon driven by video and data usage.

The biggest impact of this shift is that most cellular networks were designed and the towers spaced for outdoor customers and so the towers and radios are in the wrong places to best serve where the volume is greatest today. This trend is the number one driver of micro cell sites that are aimed at relieving congestion for specific locations.

Network Problems Can be Extremely Local. The vagaries of wireless delivery mean that there can be network congestion at a location but no network issues as close as 50 yards away. This makes it very hard to diagnose and fix network issues. Problems can pop up and disappear quickly. A few more large data users than normal can temporarily cripple a given cell site.

Network owners are investigating technologies that will allow customers to pick up a more distant cell site when their closest one is full. Wireless networks have always allowed for this but it’s never worked very well in practice. The carriers are looking for a more dynamic process that will find he best way to serve each customer quickly in real time.

Networks are Operating too Many Technologies. It’s not unusual to find a given cell site operating several versions of 3G and 4G and sometimes still even 2G. The average cell site carries 2.2 different technologies, provided by 1.3 different vendors.

Cellular operators are working quickly towards software defined networks that will allow them to upgrade huge numbers of cell sites to a new version of software at the same time. They are also working to separate voice and data to different frequencies making it easier to handle each separately. Finally, the large cellular carriers are looking to develop and manufacture their own custom equipment to cut down on the number of vendors.

Still Too Many Failures. There are still a lot of dropped voice calls, and 80% of them are caused by mobility failures, meaning a failure of the network to handle a customer on the move. 50% of dropped data sessions are due to capacity issues.

Cellular providers are looking for the capacity to more dynamically assign radio resources on the fly at different times of the day. It’s been shown that there are software techniques that can optimize the local network and can reduce failures by as much as 25%.