The internet as we know it today was released in 1981. It was designed to support about 4 billion computers. We reached that limit on February 3rd, 2011, meaning no more computers are able to be directly connected to it. To fix this, a new version of the internet was developed called IPv6. It was released in 1998. Usage of IPv6 accounts for <1% of global internet traffic. IPv6 adoption has not matched the exponential growth associated with the IPv4 internet. This growth has instead been channeled into a workaround technology called NAT that extends the IPv4 internet. The cost of NAT is that it introduces inequality and makes the internet, a now critical economic resource, less accessible and fragile. This has a high impact in the developing world. This leads to the question: Is it possible the use of NAT will continue to grow unabated and that the IPv6 internet may never happen?
Uplink Aero is a networking innovation project focused on enabling connectivity and proximity applications to grow civil society.
We think that the internet is at a really interesting point in it’s development. Even though the internet was built for ARPA, the US military’s research arm, it was designed by academics and 1960s hippies to be a very egalitarian system. The idea that end users would pay for their connection into the system was never really considered as part of the original design of the internet. The net result is that today's internet can be characterized as an arterial system, with a few (ISPs) having high bandwidth and many having less connectivity. A direct consequence of this model is that traffic peering today is predominantly a manual, rather than automatic process. This leads to inefficiencies, for example internet traffic can and routinely does flow around the world to go across the street. To extend the metaphor, traffic at the edge of the internet can be characterised as unnecessarily capillary like.
Up until recently the ISP model of selling internet access has worked reasonably well at spreading the internet in the developed world, but probably not as well at spreading the internet in the developing world. We believe that the rapidly decreasing cost and increasing density of computing in our lives now offers an opportunity to move to a free or mostly free internet. We think that under typical usage density scenarios, it should be possible for the internet to be free for end users a significant portion of the time and that the move to IPv6 could be the catalyst for this change.
The capacity of the today's internet was thought at the time by Dr. Vint Cerf, the creator of the modern internet, to be an adequate amount. But our world needs many more IP addresses, which are like phone numbers. So there is a new version of the internet. This new version, called IPv6 supports a much bigger address space of 2^128 addresses or about 340 undecillion, which is a very large number.
An interesting property of this change is that a 128 bit number (IPv6 address) is a sufficient size to encode a latitude, longitude, altitude and still have more than >32 bits (larger than the size of the entire current internet) reserved to differentiate between devices that are very close to each other in physical proximity. We think this is significant because it provides the opportunity for a routing protocol optimisations that exploit the physical topology of the earth's surface to maximize flow throughout the network.
Everyone deserves the right to an education and to lead a fulfilling life.
In December 2013, we released an IndieGogo campaign for an embedded device that anyone can run to connect edge devices to the internet, and peer traffic. To peer traffic means to act as an ISP would, and exchange network traffic with others. A key issue in doing something like this requires creating a critical mass. That involves bootstrapping a credible network. We felt we could do this by letting neighbours share internet, and tunneling our network over the IPv4 internet, routing to other UPLINK geographical networks over that medium. If the UPLINK network grew around you in physical proximity, you would have been able to unplug your pay-for connection to the IPv4 internet, and still retain global connectivity. But it turns out our marketing campaign hasn't work out how we'd hoped. So because of this we are restructuring our campaign around the project's software, rather than hardware goals. We think that in the near future device owners will also be able to extract economic value by selling their local computing resources on an open market. This would bring system level efficiencies, and lower energy usage and costs worldwide. We think the economic surplus from such a model could assist in connecting the internet to the developing world. See Eric's essay for more about this.