The modules will need to pass EMC Compliance testing and be issued Compliance Certificates (expensive!).
A suitable Cell frequency plan will need to be created, approved by the local government, and then (expensive!) Radio Licenses issued.
Licensing as a Carrier and approval to connect with the local Telcom provider will need to be arranged, along with the (expensive!) registration fees.
Then when you have it running you'll suddenly run into interference problems from all the unlicensed boxes already installed by well meaning NGO's and private organisations. Not to mention clashes with networks in neighboring countries.
Bottom line: It's not as simple as just mounting radio box on a post and switching it on.
Open Source phone towers are not a new concept. Has been already done in a number of third world and island states. But without government licensing and approval to back you up, a bunch of soldiers will arrive and simply rip out your gear.
The above questions/comments are based on actual field experience in third world countries...
You're right, but if this box is cheap enough, and we're talking about deploying 1 per village with natural spacial reuse, it's possible to see a political campaign allowing villages to install their own basestation on some empty bands.
Also maybe it could work in a city, where neighborhoods join together to request a license from the city and the city will work with neighborhoods choosing suitable locations as to not interfere too much with each other.
But of course it requires a political campaign and this might be a trigger to start that, with facebook also helping.
The lesson learned over and over in the West, is that any add-hock development of the spectrum results in huge future inefficiencies with large blocks of spectrum locked out of re-use due to poor planning in the early days.
The temptation is to let companies use new frequencies wherever they want, but in the longer term this results in chaos, and the only solution is to rip it all up and start again (at huge cost and disruption to service).
The only workable approach is to design an interlocking network of spacial and frequency allocations and then stick with it, and make sure that each block is fully utilised before moving on to a new block.
With all of the above considerations, the cost of the actual network hardware is negligible, especially if the consequence is that poor villages will have to abandon their total investment in equipment at some arbitrary point in the future.
That's true. But on the other hand, there are new techniques to let the network manage the the spectrum allocation dynamically, on it's own like "dynamic spectrum sharing"[1].
And it would be probably much easier to try such technologies on an Indian village(low spectral use, low buildings offering little interference, relatively cheaper labor).
>But without government licensing and approval to back you up, a bunch of soldiers will arrive and simply rip out your gear.
This is the single most important obstacle here. Moreover, the bureaucratic procedures one has to jump through, to get such a thing approved, are extremely opaque.
In a country like India, where the government talks about stuff like "power to the people" all the time, when someone tries to implement such a network on the ground because it's really needed and helps fulfill the ruling party's election promises, they would be stopped by local authorities who would cite some vague law or seek some hefty bribe.
Either way, I appreciate that a company like facebook would spend so much time and money to do a nice thing like this.
Honest question, what are the purposes of these licenses/fees other than to take money from citizens? If a village in the middle of nowhere builds this infrastructure to serve their own community what could go wrong without licensing?
The spectrum is a public good, and it is scarce. License fees benefit the public.
Of course, content providers and infrastructure providers do not buy licenses altruistically.
I think we're seeing Facebook being willing to invest in tech and potentially license fees because the ROI of having Facebook in a village can be estimated and offering this hardware looks like a worthwhile investment.
While this sounds like a cynical perspective, it is not. Suppose a typical company would analyze the ROI and conclude that if the payout doesn't happen in 2 years it is not worth it.
Facebook may be considering a 10 year ROI and thus may be acting both selfishly and also taking on infrastructure risk that has suffered from market failure until 2016 since no providers of first world infrastructure have opted to invest.
In a capitalist utopia, anything with positive ROI would receive investment, and all risks could be accurately hedged, allowing all available capital to be fully utilized.
As __jal points out, Ultra-wideband does not solve the problem because while it may in some cases be useable alongside existing allocations, it does not ultimately increase the information capacity of a particular piece of spectrum.
There may be some circumstances where it can help, but it does not eliminate scarcity.
Scarcity by itself doesn't seem a significant enough aspect to warrant the FCC regulation that we have though. Think of other scarce resources like platinum. If ultra-wide band (or similar) technology was used back in the 1920's, I don't think the push for regulation would have been as strong, since jamming / interference wouldn't have been a driver. I think we've got a technical solution to the original social problem, it is just that it arrived 100 years later.
Scarcity in physical objects is rather different than spectrum. For one, market demand encourages things like recycling, more exploration, etc. for platinum. Spectrum scarcity encourages more efficient use, but until we get our promised physics-violation-devices, can't make/find more.
With spectrum, there's also a real-time use component. People who consume platinum generally don't consume it in a way that in any way resembles communication use. This is related to another key reason governments hold on to the control of spectrum allocation: military interests.
>In a capitalist utopia, anything with positive ROI would receive investment
Yes. Including arbitrary licenses enforced by local governments to raise the barrier to entry and restrict interconnected innovation through forcing all players to route through a central (likely selfish) entity. This does not benefit the public.
You are referring to license monopoly. There are actually three kinds of local monopolies: margin monopoly, limited-space monopoly and license monopoly. Von Mises does a very good survey of these types of monopolies, monopoly prices and the conditions under which each exists [1]. Von Mises is a very deep thinker on this topic from a classical liberal perspective. He does agree that monopolies exist, but takes care to define the conditions under which they can exist. If a government wants to provide the best competitive environment, there are steps that can be taken to mitigate monopoly conditions from arising.
> Yes. Including arbitrary licenses enforced by local governments
Not sure how a corrupt government is part of your definition of a capitalist utopia.
In my definition of a capitalist utopia, there exists the optimal amount of regulation to maximize long term welfare. This means no regulations that benefit the next 3 years at the expense of the future, and also none that benefit the next 20 at the expense of the longer-term future. Regulations that do exist are consistently enforced without corruption or bureaucratic waste.
In our actual (non-utopian) system, most regulations have the side-effect of benefitting one constituency or interest group over another, or one relevant time-horizon over another (most are profoundly short-sighted).
Also, in a capitalist utopia there would be administrative districts that would allow for various regulatory schemes to be tested and compared scientifically, so even if one local government did make the mistake of offering a state-sponsored monopoly, the mistake would quickly be realized and corrected.
> most regulations have the side-effect of benefitting one constituency or interest group over another
Well, yes. One of the functions of government is to adjudicate incompatible desires.
This doesn't mean it always should, but something is going to play that role, and if it isn't a government, it is something that will be called a government once the shooting stops.
"Servants, labourers and workmen of different kinds, make up the far greater part of every great political society. But what improves the circumstances of the greater part can never be regarded as an inconvenience to the whole. No society can surely be flourishing and happy, of which the far greater part of the members are poor and miserable. It is but equity, besides, that they who feed, clothe, and lodge the whole body of the people, should have such a share of the produce of their own labour as to be themselves tolerably well fed, clothed and lodged." - Adam Smith
Admittedly since children cannot vote our government will always be biased toward the interests of adults and will operate on a time horizon that steals from the future to benefit the present.
However in a utopian version, perhaps there would be some mechanism by which the longer-term interests of children could be represented democratically.
I would also argue that in a capitalist utopia there would not be things that we typically call market failures, since institutions would emerge to solve the problems based on market incentives.
Also, I'm not arguing that a capitalist utopia could exist, just that it is sometimes useful to offer contrast to the idea that big firms must somehow have an enlightened and altruistic vision in order to make investments like the one in the article. It could in fact just be the true north of the profit motive animating longer-term planning than we typically see (due to perverse incentives that focus most businesses on very short-term planning).
Primarily, the fees are 1) a supply-demand response to spectrum contention (spectrum space ideal for use by mobile services is highly competitive in developed countries), and 2) go to fund expensive compliance regimes, like the expense of certification testing. This is required by industry and regulatory bodies to ensure that one manufacturer's lax quality does not adversely impact devices made by other manufacturers, which can unfortunately easily be a problem with modern radio.
At the end of the day, it's a response to the tragedy of the commons problem, designed on the theory of making the commons a lot less common (exclusively licensing sections of spectrum to approved users).
My naive understanding is it's to pay for enforcement of interference. Not sure how far LTE bands go but UHF/VHF will propagate ~80mi radius if placed on a high enough vantage point.
One of the myths is that spectrum "in the middle of nowhere" is unused.
Most third world countries have enthusiastically adoped mobile phones and have leap-frogged the creation of landline networks.
The result of this is two-fold: The Radio Spectrum is heavily used, and also very inefficiently used as so much traffic which should be carried by landlines is forced onto mobile networks.
If you've traveled in remote areas of Thailand, Cambodia or even Laos, you'll be aware of the incredible penetration of mobile phones and even the Internet.
Has there been any research into OpenBTS, Osmocom, OpenLTE ?
This is really an opportunity for Facebook to show that they are serious with their efforts to bring Internet to developing regions.
Especially with the delayed adoption of Cellular tech generations on the cellphone side, I think an open source Stack is the only way to go.
If the networks are built using decommissioned tech from somewhere else, it will be brittle and won't be cost effective in the long run, especially considering energy consumption.
The only missing part now is a baseband implementation with GSM/UMTS/(LTE) support.
Maybe they learnt their lesson from the whole Free Basics debacle and are trying to bring low cost internet to developing areas. But that's just me being optimistic. Realistic me feels like they might just use this as another trojan horse to bring Facebook rather than the internet to more people.
Exactly. Just as Uber only hires human drivers as a means of bootstrapping to self-driving Teslas, Facebook only expands Internet access as a means of bootstrapping to more of itself.
I really hate to break this to you, but that is what they (those people in developing regions) actually want. You may think you know what they want, or perhaps you think you know what they should want, but traffic stats don't lie. For people who have limited or no connectivity this is a far better option than waiting around for someone to deliver ponies and rainbows that never seem to arrive. If you wanted to start spending your time and money shipping them the ponies you think they really want I wish you luck, but they will be quite happy to just use Facebook in the interim while waiting for your magic ponies...
Fast forward 20 years when the people of those regions are unable to create an open and free internet because Facebook holds the legal monopoly in many of those areas (think Comcast)
The more that I think about this, the crazier it seems:
The argument seems to be that the cost of phone base stations is holding back development in 3rd world countries, but no evidence for this is offered.
Firstly, the world-wide mobile phone market is huge and fiercely competitive. You can now buy a 2G Chinese base station surprisingly cheaply.
Secondly, there are thousands of tons of secondhand gear lying idle. An example: Australia is about to rip out all of its 2G network. This will mean umpteen thousands of GSM/GPRS base stations being dumped for scrap.
Thirdly, the cost of the radio gear is a small part of the total cost of a base station. The cost of the Shed, Power Supply, Solar Panels and Batteries, Mast and footings, Coax, Diplexers, Antennas, etc, will be far more. Just the expense of running power to the top of the hill will likely exceed that of the electronics.
But the big cost will be extending a dozen or so outside lines to the national telephone network. Just setting up the billing arrangements, etc, will be a huge job.
The bit that really worries me would be staffing. I can imaging a bunch of enthusiastic western hobbyists turning up to install the gear, then departing leaving no-one to maintain it. I'm betting that after a week or three the whole system will be down.
But lastly is the assumption that the locals can't do all this. In any remote community there will be skilled local entrepreneurs who are selling and repairing electronic equipment. It is these people who need to be the ones who install a phone system. If they haven't yet, it will not be because of lack of cheap gear, but because of the non-existence of basic infrastructure, eg a national phone network to connect to.
And you can be sure that if there is no local phone tower, there will already be a VHF mobile radio network used by police and businessmen (eg the "red radio" VHF CBs in remote Thailand).
"In many cellular network deployments, the cost of the civil and supporting infrastructure (land, tower, security, power, and backhaul) is often much greater than the cost of the cellular access point itself."
Indeed, this is not an order of magnitude question - several orders of magnitude.
"One of our goals was to make architectural and design improvements that would result in lower costs associated with the civil and supporting infrastructure."
Hmm, really? Facebook don't think that the established players in the market are not aggressively pursuing the same design improvements?
Will this spur them on? Marginally, yes.
This is not a hard market to get into, with commodity off the shelf chips for cells being available (e.g. from Intel via MindSpeed via PicoChip).
It's good to see Facebook investing in fundamental R&D. I just can't shake the feeling they are struggling to find areas to make strategic plays if this is the best that they can do.
> The argument seems to be that the cost of phone base stations is holding back development in 3rd world countries
I'd say not quite: Since most of the first world dollars go into cutting edge technology, there is low-hanging fruit in the now-miniaturized, super-inexpensive chips that make it orders of magnitude cheaper/simpler to create reliable 2G infrastructure. So much cheaper that using second hand gear is not a better option at scale.
> You can now buy a 2G Chinese base station surprisingly cheaply.
I imagine FB engineered its own version of similar technology, but with a product focus that is aligned with the company's goals.
> Just the expense of running power to the top of the hill
It seems to be designed to be mounted on a building, so I'm guessing it's not intended to replace a massive tower installation, but instead to offer a very cheap and useful, easy-to-install access point that happens to do 2G in addition to (or instead of) wifi.
> Just setting up the billing arrangements, etc, will be a huge job.
Since the users are the product, I think it will likely be free to the end user, at least at first.
> but because of the non-existence of basic infrastructure, eg a national phone network to connect to.
I think FB has plans to offer bandwidth, both as a carrot to convince local regulators to license it, and also to incentivize various service providers to help with the rollout.
There seems to be a lot of negativity in this thread. So, I'll attempt to provide a positive spin.
I think of this (and other open source hardware and software projects) similar to how I think of mathematic theorems or tools at Home Depot. These are good building blocks to learn, prototype and build. If you want to take it commercial, then you have to bear licensing costs etc. but it is a great way for hobbyists to build things on top of these blocks rather than start from scratch.
Well the trouble with open-sourcing incomplete or very immature solutions out of the gate is that they'll get forked because people will see a need to fix what they see as fundamental flaws with an as yet unvetted or incomplete design. You'll also have people implement in production solutions based on the half-baked designs that may not be forward compatible with the ultimate 1.0 release. In either scenario you've fractured your community and diluted your effort.
By announcing a solution before it's released you can attract the attention of SMEs who might evaluate the solution and provide valuable insights and offer you the opportunity to resolve concerns before users start adopting your solution.
That's more of an open source problem than a free software problem. I recognize the concern as valid in business, but ideally it would not be relevant.
I'm not sure I understand how that distinction changes anything. You outright dismissed the project and any of it's potential merits on the grounds that it only promised to be FOSS eventually instead of initially.
Ideals are great to have but you can't dismiss something of merit simply because it's less than ideal. Ideally all medicine would be free and open source but the fact that they're not doesn't mean we should be turning our backs on them.
Considering it's not being sold this very second I'm not sure what your point is. The vast majority of open source products start out as closed sourced prototypes / research projects. Sure look at any negatives that the product may have and complain but complaining about a prototype not yet available that promises to be open source but hasn't opened sourced before being available just seems so crass and ultimately unhelpful in any respect.
This is just amazing stuff. With this, Facebook is going to revolutionize telecom in developing countries. Villagers will be able to get together and run 'bandit' networks and provide coverage to all. Laws dont apply much to startups like facebook and Airbnb, and liability is definitely not on FB.
Did anyone use the BRCK from Ushahidi? I think it had some similar goals and features. It would be interesting to compare these and see how Facebook have advanced the state-of-the-art.
Ok, let me switch off sarcasm and put it differently:
Facebook is to OpenCellular what Google is to Android: they do not really care for the disconnected, they just want a bigger market and are willing to invest in infra rather than waiting for telcos to do it.
Whose frequency spectrum does OpenCellular use? I took a quick peak at Telecom Infra Project (TIP) members page, looks like T-Mobile is participating. So can we assume OpenCellular access points RF uses T-Mobile's spectrum?
The modules will need to pass EMC Compliance testing and be issued Compliance Certificates (expensive!).
A suitable Cell frequency plan will need to be created, approved by the local government, and then (expensive!) Radio Licenses issued.
Licensing as a Carrier and approval to connect with the local Telcom provider will need to be arranged, along with the (expensive!) registration fees.
Then when you have it running you'll suddenly run into interference problems from all the unlicensed boxes already installed by well meaning NGO's and private organisations. Not to mention clashes with networks in neighboring countries.
Bottom line: It's not as simple as just mounting radio box on a post and switching it on.
Open Source phone towers are not a new concept. Has been already done in a number of third world and island states. But without government licensing and approval to back you up, a bunch of soldiers will arrive and simply rip out your gear.
The above questions/comments are based on actual field experience in third world countries...