Will Facebook's OpenCellular Project Change the World?

By Mika Skarp

It has been said that our telephone network represents the largest machine on the planet. To be sure, telecom networks are extremely complex, (count those acronyms), and are extremely capital-intensive to build and maintain. Needless to say, realizing such a machine has required a lot of co-operation and compromise between the various stakeholders, but especially operators, equipment manufacturers and governments. On the financial side of the equation this grand machine has generated heaps of money for governments by way of frequency spectrum auctions.

In direct proportion to those costs and complexity is the sheer size of most tier 1 operators, who are, in most cases, the largest companies in their respective countries, employing hundreds of thousands of employees.

This is the status quo, and it’s perfect, right?

But let’s imagine for a minute that there was an alternative way to provide telecom services. What might that look like, and what might it be? These questions came into full focus for me when I saw this post from Facebook about their new telecom node project, OpenCellular.

In a nutshell, the OpenCellular idea is to provide an open source technological platform that will allow anybody to build and power up a mobile network. This is not as radical an idea as it might seem. In fact, the core virtualized network elements of the solution have been around for over decade. The only hardware element requiring custom development is the RF front end. To be sure, developing a low-cost, high-performance RF transmitter is hard, but it is most certainly doable. So from a hardware perspective there is no real barrier to the DIY mobile network builder.

Software is defined by standards. If you have the resources Facebook does, you can easily write the code needed to deliver your desired functionality using existing standards. To your advantage (and the operator’s disadvantage), you have no legacy requirements in place that might limit or complicate your solution. In theory at least, and with these tools, you should should up with a good, clean build creating functionality in a virtual environment to drive general-purpose hardware. Another big advantage of “going virtual” is that over several generations it will be easier to fix bugs in your ASICs (application-specific integrated circuit). This may compromise performance, but your solution should provide adequately for most needs.

What are the challenges?

Though we’re pretty excited about the simplicity of this DIY mobile network project, it’s important to understand some of the issues that you, as a newly-minted mobile network operator, will face.

Hint: they’re some of the same issues large mobile operators face.

Firstly, any functionality that you might dream of adding to your network (including Cloudstreet!), has likely been patented and gross licensed. This means that Facebook needs to cover the royalties (read: significant royalties) that come with single box taken in to use. And in the spirit of open innovation, the standards defining these functionalities ensure that future advancement will be granted on a fair ground, preventing old players from preventing new players entering the market. Still, taken as a whole, the project shouldn't be completely cost-prohibitve to implement.

But let’s take a closer look at the model to see what makes it tic.

Right out of the gate, and in order for the whole thing to work, Facebook will need to hand out (or more likely sell) the boxes to this brave new generation of DIY operators. This is the most difficult part of the plan, but it isn’t the part that might make it totally impossible in some jurisdictions. Keep reading.

Of course, all mobile networks are and must be regulated. In order to run a network you need to acquire a frequency license from the local government authority. Easy right? Yeah, not so much. This factor alone would likely preclude Facebook from operating networks in many markets around the world (including Europe) as they would need to negotiate with each national authority, and quickly get bogged down in the sheer technocratic quicksand of it all.  So the short answer is, in many cases “it simply may not fly”.

That said, ITU (Telecommunications Standardization Sector) should open unlicensed spectrum for DIY and small operator mobile technology, like they do for WLANs. Ground-sourced mobile network would certainly be a cool thing, but to make this happen will take some time and co-operation for sure.

What's the Driver?

So apart from geeking out on mobile nodes why would anyone play? And further, if it’s going to be both expensive and likely impossible in giant swaths of the planet why is Facebook’s open source mobile network so important today?

As people in the developing world know, there is a huge digital gap between places where Internet is available and where it is not.  Back in 2014, Mark Zuckerberg put the challenge to us in this way:“

"The internet is essential to growing the knowledge we have and sharing it with each other. And for many of us, it’s a huge part of our everyday lives. But most of the world does not have access to the internet. Internet.org is a Facebook-led initiative with the goal of bringing internet access and the benefits of connectivity to the two-thirds of the world that doesn‘t have them.
Imagine the difference an accurate weather report could make for a farmer planting crops, or the power of an encyclopedia for a child without textbooks. Now, imagine what they could contribute when the world can hear their voices. The more we connect, the better it gets."


There is clearly a very big business interest for Facebook to bring the Internet to every human being on the planet, but more importantly it is a question of human rights, which include open access to the Internet. If there is a credible threat to carriers in the concept of unlicensed frequency, they should take the long view. It will only accelerate modern technology roll-outs to places where they currently have no reach.

A good analogy for this issue is the struggle with patent holding pharmaceutical companies over the accelerated arrival of lower cost generic alternatives. This very quickly becomes an ethical issue that corporate enterprise must work to find itself on the correct side of. And in this case, regardless if these roll-outs are based on Facebook technology or not, they will certainly benefit innovation, the enterprise, humanity and Facebook (not necessarily in that order).

Could Facebook's model work in western countries?

There is no question that technically it is a brilliant development, and from a cost perspective it may well be feasible. It brings the concept of a repeater to a whole new level of intelligence, while being considerably less complicated and costly to implement. And while governments and standards bodies might be open to making frequencies available for the DIY set, the question arises: will operators be able to take advantage of the box as well? There's no question that operators will find the technology handy in places where coverage is poor,; in subways, parking garages, basements, bunkers etc. If this does take hold it will ,no surprise, happen first in the US and quite likely never in Europe...sigh.

To conclude, I am happy to see the telecom world welcome a new competitor in Facebook. Joining the last major entrant Huawei 20 years ago we see this as a potential game-changer on a major scale that will make some people some money, but more importantly may make the world a safer, happier, healthier place.