Flipping the Internet upside-down.

I think most big decisions are better made on the basis of bottom-up, democratic voting principles. For that reason, I’ve recently started to see the Internet as fundamentally flawed. This flaw is not in the design, but in the implementation — the steps it has taken between the lab and the “open” market. I see the current Internet as an institutional, top-down, dictatorial development, that has enabled (yet so far suppressed) distributed deployment. It started with large Military and University institutions, with the intent of better distributing reporting mechanisms, for transfer back to a central authority: DARPA. The inter-institutional mesh they designed had an amazingly robust design for war-crisis routing, but it tended to devolve into a more vulnerable hub-and-spoke distribution within each institution.

The grant of Internet technology to private industry followed a similar pattern. All roads onto this new civilian network lead through our existing regional communications market operators: the “Baby Bell” telecom incumbents. Even the secondary commercial channel goes through an existing central-broadcast incumbency: the local television cable operators. Both institutions follow the same hub-and-spoke distribution model within their “last mile.” The core hub represents the authority, the incumbents, while the spoke ends represent us, the little people — the “subscribers.” There is currently no such thing as an eCitizen — we can’t even vote on our Internet connections! We’re all just “subscribers” paying for a “service,” no matter how few providers actually exist, no matter how powerless we are against them.

What we need now is a new, bottom-up, “first mile” version of the Internet. This new vision for the Internet can exploit these existing institutions, yet it does not require any of them. I think we can build something even better than Network Neutrality into this new design. Instead, we can build a people-centered network. The only prejudices we should have to deal with on this network are our own.

The top-down nature of the Internet today is reflected in everything from the basic command-and-control reporting nature of Transmission Control Protocol (TCP), up to the central authority represented by the Internet Corporation for Assigned Names and Numbers (ICANN). ICANN rules over the current Internet in the form of proprietary allocation methods for Internet Protocol version 4 (IPv4) numbers, and for domain names (like the one you see in the address bar of your web browser). ICANN IP number block grants also carry an implicit expectation of pre-defined subnet hierarchies within each Internet Service Provider (ISP). Even commercial Local Area Network (LAN) hardware works largely on hub-and-spoke deployment assumptions.

Via a little smart design, and a lot of luck, Internet Protocols (IP) are still amenable to a more distributed, bottom-up network designs. User Datagram Protocol (UDP) is more like a simple envelope standard than a TCP postal officer. Internet Protocol version 6 (IPv6) can provide ample addresses to form a foundation for an institution-less (in my way of thinking, user-centric) network design. A single IPv6 64-bit subnet provides ample space to geographically address more than one unique device on each and every square centimeter of our Earth’s surface. Geographic and ad-hoc routing methods enable efficient traversal with inexpensive “sight limited” routing hardware. You don’t have to memorize a map of the world to tell someone how to get to Main Street and 5th, so neither should your IP router. With such a design basis, talking to your neighbors over IP should be easy. Talking to the whole world over IP becomes a side-benefit, not a core operational requirement.

The top-down nature of the Internet today is reflected and enforced by the analogies we use. Network connections are not like “pipes,” where resources come down from some central reservoir. They are not like electrical “lines” either, where energy is created and distributed from a set of far-away generation plants. Modern networks aren’t even like old telephone connections, where some central switching-station routes all regional calls. IP packet networks are much more like the “roads” we navigate every day. Everyone should be able to have a driveway connection from their personal property onto the shared road system. Everyone can provide their own “vehicle” (device) for utilizing these roads. Every “trip” has a unique start and destination. Vehicles obey regulatory requirements, and drivers obey the rules of the road, yet everyone is considered to be “free” in their travel plans and methods.

Unlike paved roads for automobiles, wired roads for data are only a few millimeters wide and very flexible. Very little land or resources are required to install data roads — a small fence-line conduit is sufficient. Unlike vehicle traffic corners, the traffic signs and lights of a data road intersection can be handled in a small box, taking very little power, all procured and installed for insignificant cost. Unlike airplane travel, data travel in the air can be launched from a small roof space. No data pre-flight schedules are necessary — just hop in the air when it’s clear! These data roads, traffic intersections, and air ports can all be installed and configured for far less than the price of a small car. We still don’t have flying cars yet, but we have plenty of invisible flying data!

Small home and land owners can only own their vehicles, garages, and driveways onto transportation roads. With data roads, land owners and developers can build an entire set of streets and intersections, and operate independent airports, all connected directly to friends and neighbors. A complete set of data travel infrastructure can be encapsulated within a single household. Nobody has to go all the way to the highway and back, nor pay any highway toll, just to ask a small favor from a neighbor.

A sense of responsibility and ownership can extend from more than just a communications device — we each can all own our small piece of the larger inter-networked whole.