Connectivity for the unconnected: a double bind

During the Mobile World Congress in Barcelona this past February, Facebook made an “historic” announcement that they were launching the Telecom Infra Project (TIP) in conjunction with a number of large telecommunications companies and manufacturers. The new initiative takes a page from another Facebook-led industry collaboration called the Open Compute Project [] and is all about vastly lowering the costs for telecommunications carriers to build new networks by creating inexpensive, open-source software and hardware. This level of collaboration between Facebook and major telecoms companies is unprecedented and signals a shift from what has traditionally been an adversarial relationship, at least publicly.

So what is this about? Why the collaboration all of a sudden? The answer is fairly simple. This initiative is about getting more people on-line by making it less expensive for traditional providers to build out their networks into places without coverage. Something like 6 out of every 10 people with access to the Internet use Facebook. So the company must focus not only on trying to turn 6 into 7 (growth in existing markets), but also on expanding the total number of people on-line (creation of new markets). An anecdote that I find useful here is to imagine a publishing house so successful that its main obstacle to selling more books is illiteracy. TIP is about Facebook finding ways to teach people to read. Companies like Google and Facebook simply need to get more people on-line and they automatically make money. This is why they are underwriting network rollouts and innovation, through project Link from Google and the TIP project that FB is coordinating. For the major telecoms companies the benefit of a project like TIP is that its stated aim is to make it easier and less expensive to set up new networks.

These two initiatives, Link and TIP are more grounded (literally) elements of both companies’ access and connectivity portfolios that are otherwise fairly cynical attempts to connect people to the Internet without actually involving them in the process. Think lasers, balloons, drones and so on. There has also been a quite a kerfuffle internationally about the Free Basics platform that Facebook has been fitfully rolling out, and prior to that around net-neutrality. Internet access for those who don’t have it “yet” and how to go about it is an issue with many facets and competing interests.

Talking to people who move bits from one place to another in Mexico (small ISPs, mainly), they claim that around 80% of the traffic their networks transport is related to Facebook, Google, and Microsoft. This is troubling in that, even with open standards and net neutrality in place, the web has been conquered to such a major extent by large companies. Perhaps, in some round-about way, it does make sense for these big companies to underwrite the construction of new networks. Yet the proposition is a scary one. Imagine the control these companies have and continue to amass when they not only control the content platforms but the physical infrastructure over which data is transported as well.

When we talk about access or connectivity, about getting more people on-line, what are we really talking about and what can we expect? As time goes on, the “Internet” has become more and more about serving content to users than about peer-to-peer sharing, as in the salad days of the early, more peer-to-peer internet. As some have argued, the internet is increasingly becoming like TV – a tool of mass media – yet more insidious. Early theorists about the development possibilities of mass media are not wrong in saying that TV or radio is useful for this purpose. But as those media have shown, and the internet as a new form of networked media is showing, much of what mass media delivers is mediocre at best and perhaps even culturally damaging. The negative impact of these media is more palpably disruptive to people from non-western cultures, who very rarely get a hand in creating or distributing their own content in their own languages. In many cases, especially in countries where technology is imposed as part of foreign aid or wholesale reform packages, information and communication technologies act as vehicles for the rationalization of society through the introduction and bolstering of market logic. In this sense, the mere existence of mass media and networks lays the cultural and economic groundwork for their own perpetuation.

So while access to information is a good thing, what is the baggage connectivity brings along with it? Probably the most harmful is the opening up of marginalized and generally “non-essential”, peripheral people and places to data mining and surveillance. The economic value of reaching new populations is difficult to quantify but certainly substantial. Economics aside, though, there is an undeniable geopolitical value for those in power in aggregating and analyzing behavioral data in the poorer parts of the globe where unrest is most common and likely.

In a broader context, telecommunications services are treated as a good or service, and so ensuring these goods are offered in the market is of utmost importance as this generates revenue directly for the companies that provide the services to customers. Secondarily, as these services or goods exist within a market-based global system it is also worth considering their usefulness in facilitating international commerce. In fact, one might argue convincingly that our current world market could not exist in any recognizable form without digital communications infrastructure. In this sense, telecommunications have become – and perhaps always were – a strategic sector with regards to the furtherance and reproduction of global capitalism. And so the “system”, as it were, ensures that these networks exist and reach where necessary.

There is a strong argument to make that a third way in which telecoms underpin global capitalism is through the facilitation and expansion of state surveillance and corporate data-mining. Over the past couple decades we have witnessed the construction of an unprecedented surveillance apparatus that threatens the fundamental rights of all people, and that creates an environment of panoptic repression and control. This is good for those in power who want to stay there as well as for gathering consumer information and feeding increasingly “intelligent” corporate algorithms.

With respect to expanding communications in rural areas, there has historically been a tension in that it is not economically viable to provide telecommunications services with an eye towards profit in many places on earth and yet, as argued above, for reasons of national security and the growth of capitalism, it is necessary to incorporate ever larger areas and populations. The special treatment of telecommunications is evidenced by the nearly endless litany of studies and news items proclaiming that there are more people with mobile phones than access to clean drinking water or sanitation facilities, and that network infrastructure is already more pervasive than basic infrastructure (electricity, roads, housing, etc) in many places. Something seems off, when those on the margins can’t find a job or shit in a flush-toilet, but can access Facebook from their cell phone. The point here is not to moralize, but to understand why this is happening.

Traditionally, big telco companies pass the baton to governments in places where there is no direct profit motive and the latter steps in with digital inclusion strategies (many of which fail) and universal service funds (which are also abject failures). To wit, the US State Department started something called Global Connect to bring an additional 1.5 billion people online in developing countries. In Mexico, there is an office called the Coordination of the Information and Knowledge Society, who’s stated goal is: “Driving effectively the country’s transition to the Information and Knowledge Society, integrating the efforts of various public and private actors in this task, and drawing on all Mexicans to join this process”.

Essentially this is an example of how the State is responsible for ensuring the conditions for the expansion and deepening of the logic and economic model of neoliberal capitalism, which, if history is any indication, is not necessarily something people in rural areas want or will find beneficial. And now the aforementioned baton is being passed back to corporations, albeit of a different ilk as big internet companies like Google and Facebook are proclaiming their interest in dealing with the problem of rural connectivity, and are enlisting the help of large telcos. This troika of Big Internet, Big Telco and the State might just get internet to everyone on earth, but at what cost and in whose name? Unless we radically re-imagine networks: how they are built, managed and how people engage with them, there seems to be no way that pursuing increased connectivity doesn’t also do a whole lot of harm both directly and indirectly to the people it is meant to help.


The case for social rights for radio spectrum

In the simplest sense, radio spectrum is a way to classify certain electromagnetic frequencies, between about 30Hz and 300Ghz, that are useful for the transmission and reception of information with currently available technology. Radio spectrum is a socially constructed idea insofar as we have created the concept in order to address our ability, through technology, to harness electromagnetic waves and transmit information over distances. We then assign this concept, Spectrum, a certain meaning that in turn dictates how society views and deals with it. In other words, radio spectrum is a social construction.

Beyond the socially constructed concept of spectrum, radio spectrum is also a socially produced good. We generally miss this as it is so often and wholly transformed into an economic good. Radio spectrum is materialized and commodified so we can think about the underlying idea more easily, so we can regulate it, and so we can rent it out for use. The general treatment and understanding of radio spectrum not withstanding, its true value derives from its social use and is ultimately worth the communicational potential it has for the users of a particular frequency or portion of the spectrum. This is important. To wit, the winning bid at a recent auction of AWS-3 spectrum in the USA was $45 billion dollars. Why did the winning company pay so much? The answer is simple: they know that they can put up infrastructure (another large investment) that relies on this spectrum to operate and that will carry the communications of millions of users, whom they will charge for doing so.

As price is governed by the laws of supply and demand, the price of spectrum, on the demand side, is ultimately determined by what users are willing to pay for telecommunications services which are dependent on particular parts of the spectrum to function. On the supply side, a major issue affecting the price of spectrum is its scarcity, whether real or artificial. In one sense, spectrum is scarce in that two parties can’t occupy the same part of the spectrum in the same place at the same time. Spectrum is currently regulated to ensure interference cannot happen as only one party has exclusive use of each frequency. This is an inefficient, out-dated approach and ultimately creates more scarcity. Yet in another sense, there is no real scarcity issue as the spectrum and the frequencies which comprise it cannot be used up or even permanently degraded as a natural resource like land or water might be. Better technology and more sensible regulation mean there is potentially plenty of spectrum to go around.

As radio technology has advanced, spectrum has increasingly undeniable and nearly limitless use value. However, through conceptualization, commodification and regulation, spectrum is then transformed into something which can, and almost always must, be paid for in order to use. Case in point: only 2% of radio spectrum in Mexico can be used without a license. Same goes in most places. Spectrum, for lack of being a material thing or a unique creation of human ingenuity (think intellectual property), can’t actually be owned in the proper sense. The social and political construction we call the radio spectrum is only something we can use. So why do we assign property rights to something that is immaterial and not the creation of an individual intellect? The answer is obvious: in order to charge rent (see above regarding the AWS-3 auction). Rent here is defined as the benefit received from assets formed by creating official privilege over natural opportunities. In almost every case, the official privilege is that of the State. But how did this come about?

Let’s use Mexico, where we do much of our work, as an example. Mexico is a sovereign nation and this sovereignty extends and applies to the national territory, including airspace, waters, coasts, etc. Through some circular logic and pretty weak legal reasoning, Mexico claims that the radio spectrum is part of the airspace of the nation, and therefore is under the dominion of the people, represented by their government (a.k.a. the State). As we’ve briefly discussed above, the radio spectrum in no way, shape or form has anything to do with the air or the airspace. The spectrum exists insofar as we have a means of transmitting and receiving information over certain frequencies. So in this case, as is so in most countries, the task of regulating the use of spectrum falls upon the State as a matter of jurisdiction. This arrangement is propped up by organizations like the International Telecommunications Union, which is the branch of the United Nations that deals with (you guessed it) telecommunications and radio spectrum.

We make radio spectrum a thing, then we allow the State, as a territorial right, to regulate it, and subsequently rent it out to the highest bidder. Under this logic, the companies that pay (generally lots of money) to use spectrum have the most self-interest in ensuring that end users can access spectrum via the services they offer and charge for. This is true to a point. One only need to look at the boom in cellular coverage around the world in the last decades. But what of those many billions of people that are not covered by commercial services? How are they meant to meaningfully communicate in a digital world if they can’t access spectrum?

There is frankly not much we can do about this whole arrangement in the near future short of unraveling decades of international treaties, national laws and billions of dollars in investment. What we can do, however, is to bring to the forefront and stay ruthlessly on message about is the fact that spectrum is socially useful and should be regulated in such a way as to maximize its social usefulness for all. Not just for those in cities where commercial services exist, but everywhere people wish to communicate over distance. I would also argue that we should attempt to have spectrum considered as a human right, and specifically an economic and social right. The right to freedom of expression and the right to information are codified as human rights at pretty much every level. Economic and social rights, a subset of human rights, exist to guarantee that every person be afforded conditions under which they are able to meet their needs. If communication is a human right, how do we then ensure that the ways in which we communicate are also protected as such and that access is possible for all? In other words, there should be a right to spectrum just like there is a social and economic right to housing or education.

Why sites go down

Rhizomatica has recently moved into a phase of critical reflection about what we do. Specifically, we are looking to share not only our successes, but also the challenges we face, how the affect us and what we are trying to do about them. The idea and practice of local or alternative networks is not new, but is a space where we think there need to be some hard conversations about how to expand them and make them more resilient. This starts with being honest about our own shortcomings and the difficulties that exist when interfacing with the traditional or corporate infrastructure that we rely on to move some of our data or traffic around. So with that in mind, an explanation of why some sites die or go down for longer than a few days. Please note these are not in any particular order.


Reason 1. Political or Social problems

Rhizomatica has decided, for a host of reasons that are beyond the scope of this post, to work directly with organized communities and social organizations, rather than a more entrepreneurial or private investment approach. This means that while we are protected from some things such as vandalism (much more likely someone will steal something belonging to someone else than to themselves) it puts the project at the whim of local political issues. We have one site that has been down for 2 or 3 months due to a disagreement between two political factions within the town. This has meant that the administration computer was held “hostage” and no users could pay their monthly fee, buy credit, etc. and that the town stopped paying it’s Internet bill and therefore had no long-distance service. While the places we work are generally quite organized, these types of political disputes do happen and are pretty much impossible to do anything about from our end. Mainly we just have to wait for things to work themselves out locally.


Reason 2. Existing networks, “Interference” and Colonialism

We find that our service struggles when there is already a corporate telco network present or one comes along in a place where there is a community network. We’ve found there are a few explanations for this. The first is an issue around how people experience their phone and network and are conditioned around what traditional providers provide. The second is something more ephemeral relating to how people view development, modernity, progress and so on. This mentality is in turn shaped by the history of colonialism in Mexico. What this all means for our work is that we face, in some localities, expectations of what the community cell phone service should be like. So even if it’s vastly cheaper and more or less of the same quality, the fact that we don’t have direct interconnect and users have to dial using the country code or folks calling in hit a virtual exchange and then need to enter an extension to reach their destination, makes it so that some people prefer to use the more expensive, corporate network. Sometimes its just pure brand recognition. For some rural towns being able to show off that you can afford to use the corporate network holds prestige.

Another related issue we face is the existence of fixed wireless terminals; a fancy name for a cell phone that looks and operates as if it were a fixed line phone. Since coverage is scarce in many rural areas, people resort to buying these and attaching high gain directional antennas in order to pick up signals from really far away, behind mountains and so on. Strictly speaking this isn’t legal, but who can blame folks for doing what’s necessary in order to communicate? The problem for us comes in when the community installs a network and the fixed wireless terminal tries to negotiate between the community network right there and the weak signal coming from far, far away. The result is that the weak signal sometimes seems to disappear as far as the phone is concerned and therefore it can’t be used. This angers lots of people and has led to us having to shut down two networks permanently. If you understand Spanish, you can read more about it here>>


Reason 3. Infrastructure problems.

This is something that we’ve spoken about before and refers mainly to electricity and Internet. As a hybrid approach to networking, we rely on the existence of other systems to do what we do. This is certainly a vulnerability. It also makes what we offer much less expensive. To the point, we have had equipment damaged by electrical surges, or the whole system will go off when the power is out, which can sometimes last for days. The situation with the Internet is more or less the same, with the added issue that, at least in Oaxaca, the major routes to the IXP and the US seem to be constantly broken or not operating to any reasonable standard. This means it is really hard to connect long distance calls for the users. The solution to the electricity issues in most cases is having a backup battery and protecting the equipment from surges and discharges. But fixing the Internet is hard. Maybe we’ll end up trenching our own fiber to Mexico City 🙂


Reason 4. Damage from lightning

This plagues pretty much every rural network and relates partially to Reason 3, when the bolt strikes the power lines rather than the tower or installation directly. For direct strikes, we are working to find a cheap and reliable way to protect everything we put on the tower and everything that it connects to. It’ll add about 5 – 10% of the total installation cost, which is reasonable.


Some of the above can be fixed with a little more investment. Some are complex, deeper issues that we can’t fix on our own. And yet others are tied to the way we have and would like to continue doing things because they align with our values as an organization. We want to offer free/open, inexpensive communications options to people that allow them to both organize collectively and strengthen existing community institutions. We probably focused too much on getting costs down and not enough on making the networks robust. So now we can adjust and bring things into better balance.

If you haven’t seen it already, check out our friend Steve Song’s blog post about the potential of new approaches to cellular.

We are particularly excited about the fact that the new spectrum rules were published in April and we appear in the Federal Gazette equivalent as the reason for the new law.

Dealing with technical issues

We have 19 networks up and running around Oaxaca now and have decided to take a pause in regards to new installations in order to focus on tightening up our support infrastructure and processes to actually be able to attend to the issues that arise.

With all these networks running, we have zeroed in on some important points of failure.

First and foremost, the FLOSS GSM stack we use has some kinks that need to be worked out, which we are doing together with Holger Freyther from Sysmocom/Osmocom. These pertain mainly to paging problems, broken channels, and a big one that is perceived by user around how the audio is handled. Beyond that our hardware setup has been really stable in almost all cases, excepting one major ligthening strike that fried what we had in place to protect the BTS and also part of the BTS itself.

More generally, what we are seeing with regards to problems are more related to the reliability of the Internet we use and the electrical grid. From the Internet side, no matter with WISP we are working with, we are finding that there is congestion at the fiber that leaves Oaxaca, and consequently the VoIP aspect of what we do becomes impossible for some minutes or even hours every day. This issue, like the electrical grid issue, are related to major, national deficiencies. This is frustrating as it feels like there isn’t much we can do about it. We have begun a process at looking at other fiber options in Oaxaca with the government, but that is going slowly.

So much going on!

Some big pieces of news, lately.

The first one is that we just installed our 10th site. I am not sure if any of the Rhizomatica team thought we’d actually get to this point. When we started out community cellular we just an idea. There wasn’t really any afforsable hardware that worked well, the software had all kinds of bugs, and so on. We still have lots of challenges to overcome together with our partners and suppliers, but we received a huge piece of news at year’s end that seems to point to the continued existence and expansion of community cellular, at least here in Mexico.

Which brings us to the second big thing, one that has been the result of months, if not years of work by Rhizomatica, and lots of good timing. The regulator here in Mexico, IFETEL, just published it’s National Frequency Attribution Plan and:

“For the first time specific bands have been assigned for social use in the telecommunications sector. As part of this, various portions that are available within the segment known as the cellular band, between 824-849 and between 869-894 MHz, are now available for concessioning.

It is proposed that these portions of the spectrum are to be concessioned for the provision of rural connectivity, which could meet the immediate needs of basic telephone service in regions not served by existing licensees.”

This means that the entire country, minus Mexico City and the surrounding states, can now benefit from community networks like the one’s Rhizomatica helps to operate. This is huge and unprecedented, as far as we know.

And for all of you Spanish speakers, here is a video for your enjoyment.

GPRS is Go

GPRS testFirst data transmission tests for our GSM networks