Lead, Follow, or Lose the Great Game: Why We Must Choose
a US IPv6 Leader
IPv6 can be the basis either for the US to become more deeply interwoven with its military allies and key trading partners and pay for its physical imports with data flowing via our world-class end-to-end networks, or to become more isolated from the rest of the world, with less and less that other advanced nations need or want. The US has gone from being a net exporter of virtually every major category to being a net importer of food, goods, capital, and high technology in 2005, even as the dollar heads towards the currency equivalent of being de-listed as the world’s reserve currency. Federal investment in IPv4 leadership has enabled us to be a net exporter of data, media, and services, but if the US does not win the game for IPv6 leadership, we put those exports at risk as well. Must we ask, “Who lost the Internet?” in years to come?
The choice of winning or losing IPv6 is in the hands of a few hundred federal government officials who decide whether and how to collaborate with the IPv6 leaders of key US allies including Japan, Korea, and Germany, as well as multinational alliances like NATO, the European Defense Force, and the European Commission. All these powerful entities will be sending representatives to the Coalition Summit for IPv6 to meet with their counterparts from the United States federal government, and with each other. The question that the US federal government needs to answer is, “Who speaks for IPv6 in the US?”
Since June 2003, when then-CIO John Stenbit announced the DoD Mandate for IPv6 by 2008, the Dept. of Defense has been the leading voice on IPv6, and the Dept. of Defense gave by far the most comprehensive and intelligent plan for IPv6 Transition at the most recent IPv6 Summit, the United States IPv6 Summit 2004 last December. However, the Dept. of Defense has never claimed to speak for the rest of the US federal government, nor for the 50 US state governments, nor for industry or academia or standards organizations, and Congress has not yet acted to help IPv6, though IPv6 will impact everything Congress does for decades to come.
The answer to this question of “Who speaks for IPv6 in the US?” is that at the moment no one is in charge of IPv6 for the US government as a whole, from either the legislative or the executive branches of government.
Does it matter? Yes, an actual identifiable leader for IPv6 is vital for a number of reasons, including helping federal CIOs to move to IPv6, keeping comparative statistics and informing Congress, acquiring v6 addresses for the US, overseeing major tests, initiating international partnerships, responding to IPv6-related power moves, and keeping allies happy and busy working with the US. Most importantly, we need to recognize that IPv6 creates a world-sized new game board, that a game similar to an astronomically scaled up version of Othello (a game with similarities to the ancient Japanese Go) is being played, and that the US needs to get someone to play our chips for us. Otherwise, this great game will be over before we start playing.
The US federal government needs someone to lead a team to come up with a federal IPv6 transition plan, starting with an effort to reach out to the 150 or so federal agency Chief Information Officers (CIOs), and get input on what each CIO wants to do, and then give guidance on what each CIO needs to do as part of the whole federal effort. Each federal CIO needs the chance to be able to ask questions and to get clear, consistent answers and direction. A consistent definition of IPv6, certification, recommended events to attend, and more are just a few of the hundreds of questions that will be asked and need to answered by the IPv6 leader. These questions cannot be outsourced because the potential for conflict of interest is too great.
The federal government’s executive branch needs a person who will track IPv6 growth and usage in the US and in all other countries, and make a quarterly report of what countries and companies are leading the Internet, in much the same way that the Dept. of Commerce tracks imports and exports. The federal IPv6 champion should also be prepared to learn from, and to politely advise, Members of Congress and their staffs and committees, and to testify at hearings related to IPv6 and commerce, trade, defense, homeland security, standards, and communications. The federal IPv6 leader also needs to coordinate with components of the Executive Office of the President, including the Office of Management and Budget, as well as the General Accounting Office, making sure that IPv6 transition is budgeted for, and that the funds are spent optimally. If possible, the federal IPv6 leader should be able to get IPv6 leadership on the agenda of the President, who would advance US IPv6 chances immensely simply by mentioning it in a speech, as the past two prime ministers of Japan have done, thereby accelerating adoption of IPv6 in that country.
The US federal government needs someone to estimate how many IPv6 addresses are needed and to make a comprehensive and persuasive case that will go to ARIN, IANA, and probably to ICANN, to get a very, very large block of IPv6 addresses. The smaller the number following the slash, the larger the fraction of the entire 3.4 x 1038 set of IPv6 addresses. Both the size and the date of request could be plotted to give the relative ranking of a sort of Bell Curve for Countries. The greater the block of addresses, the more that can be done without having to go back and fight for more addresses in the future, especially since IP addresses are historically harder and harder to get over time.
The most foresighted companies, such as Vodafone (/20 allocated) and TeliaSonera (/20 allocated), got their IPv6 blocks months ago. Companies ahead of the curve such as Google (/32 allocated) recently got their blocks. However, other than the Dept. of Defense, none of the federal agencies seem to have requested their blocks of addresses, indicating that they have no plan to get the addresses while it’s still relatively easy to do so. This indicates a power vacuum that someone in the federal government, our IPv6 champion, needs to step into and fill, getting the address blocks on behalf of the civilian 2/3 of the federal government, and then allocating them (and then hopefully leading cheers, as different agency IPv6 transition plans are created and approved).
This group request function also allows federal agencies an additional layer of privacy, because the requests for address allocations, if done at an agency or sub-agency level, will end up revealing plans about America’s critical infrastructure that might be better kept FOUO (For Official Use Only) rather than put into the public record, especially for those agencies dealing with terrorism, law enforcement, and cybersecurity. It might be better, for instance, for the NSA, CIA, and Dept. of Homeland Security to explain their plans only to the federal IPv6 champion, rather than directly to groups that aren’t structured to deal with classified information, and would probably feel more comfortable not hearing it at all.
The US federal government needs to be responsive to the growing needs of allies that want to test IPv6 and see how it works as it scales. There are excellent companies that provide test equipment, including Spirent and Ixia (both of which provided articles for this issue of 6Sense). Such companies should be taped to be part of a very large scale project to speculate on what the IPv6 Internet will look and behave like if hundreds of millions of Americans plus hundreds of millions of Europeans plus hundreds of millions of Asians are all using IPv6 with broadband, wireless, automobiles, and appliances. Someone needs to not only plan the project (involving dozens if not hundreds of vendors, manufacturers, and integrators), analyze the data, and publish the results, but should also publicize results as a PR vehicle to get industry, domestic government agencies, foreign government agencies, academia, and the media interested in IPv6 and its progress. The IETF's 6Bone, GEANT in Europe, Internet2 and Moonv6 in the US, CERNET2 in China, and WIDE in Japan have all made good use of limited funds to research IPv6 (especially Moonv6, which did the most with the least), but we need to scale up tests to be 10 to 1,000 times larger to get a better sense of what the future will hold. Without a central federal focus or someone in charge, none of the tests will be large enough to give us the data we need to make the best planning and decisions today.
The US federal government also needs to take the initiative and be a champion of different kinds of projects that will involve our allies. China’s Academy of Science and Engineering has done a good job of reaching out to individuals, companies, and even countries to invite their participation in China’s IPv6 projects. As it does so, it also creates the network equivalent of a gravitational pull towards integration with Chinese networks, using Chinese rules and policies -- and away from the US. At the recent US IPv6 Summit I arranged a meeting with a very large telecom company owned in part by the government of a key US military ally. Representatives of this allied country’s military, telco, and national IPv6 bodies were present, along with a US government representative. They said that they had joint ventures at the government and corporate level with China, and would like to also do joint ventures with the US at government level as well. This company, out of respect for the US military, was willing to set up an entirely separate team to develop IPv6 with the US. To date, however, no joint projects with the US that were proposed by this country, or any other countries, have been officially announced, even as China’s government (via the Beijing Internet Institute and multiple companies in which officials in the Chinese government are stakeholders) has continued to create joint ventures with other countries. The risk is that, in the time it takes the US federal government to get its act together, a dozen IPv6-savvy entities tied to a major player such as China could engage in dozens of international projects, and be so far ahead in scope and scale that the US would have little to offer in the way of research. Such a group could up the ante and say, “You can work with the US, or with us, but not both. Make your choice.” Using this simple tactic, a foreign group could weaken alliances and relationships that America has built at tremendous cost in resources, including the lives of its citizens. In a world where all militaries are moving towards net-centric warfare, teaming to gain IPv6 expertise might be as highly prized as teaming to gain expertise in precision-guided munitions.
The US federal government needs to have a single voice on IPv6 to be able to respond, favorably or unfavorably, to bold proposals from both friends and rivals. In recent months, China’s ambassador has criticized ICANN on the floor of the United Nations General Assembly and said it should be under ITU control, and China’s Houlin Zhou, director of technology standards at the ITU has both affirmed this idea, and told the Chinese press that Internet addresses should be allocated based in part on population (it was not clear whether he was referring to IPv6 addresses or existing v4 addresses). It is important that someone in charge at the US federal level should actually care about Internet governance and IP address allocation, and express an official opinion, one way or another, with respect to any notion or proposal floated by an official foreign government body.
The transition to IPv6 could prove to be critically important because IPv6 is three things in one: an addressing schema that can allow everything and everyone to connect to everything else, a technology that will allow the fast to eat the slow, and a platform for partnership that can be used to woo allies away from one alliance to another. Winning what we could call The Great Game of IPv6 will be essential to winning the future, because IPv6 nodes are end-points, and the dynamics of a game of end-points are counterintuitive. You can be winning the game in points for 30 moves, and then lose the game on just the last move, and lose it badly. With IPv6, the potential exists for the US to seem to be winning in many areas, militarily, technologically, economically, and then - BOOM - be left devastated by a surprise series of moves with the result that it has clearly lost, and almost no one will have understood why because leaders and the public didn't know what game was being played. I believe that one game best models the complexity of the battle for IPv6 leadership - Othello - because Othello is about getting end points and then flipping over everything in between to be like the end-points.
I would ask any reader who has gone with me this far to go and play a half hour or so of the game Othello, available online at http://www.mattelothello.com and other game sites. In Othello, whoever has the most chips their color at the end wins the game. What is important is that players need to be very careful about edges, and especially the four corners, because a well placed corner piece could theoretically change the colors of 22 other chips in a single move – both sides and the diagonal across the board.
The analogy for the Internet lies in how Dr. Larry Roberts got people to start linking into and sharing resources with the ARPAnet. Drs. Vint Cerf and Bob Kahn got people to start using TCP/IP, and IPv4 was born. The Internet started expanding, and every network or computer caught between the end points ended up “flipping” to being IPv4 enabled. Within a decade of the invention of the router, Cisco became the greatest venture capital investment of all time because it “flipped” so much equipment in between end nodes that you wanted Cisco equipment to be on the edge of the network. I believe that if a foreign group of which the US is not a part reaches critical mass (i.e., dominance of end points) in IPv6 that it will have a huge advantage in selling its hardware, software, services and subscriptions in a winner-take-all world. Half of all enterprise transactions now flow through SAP software. Imagine a world in which 90% of all information went, end to end, through networks of cables, switches, routers, software, mobile phones, laptops, desktops, printers and IP-enabled cars that were controlled by one foreign group (which could be hostile to the US), because IPv6 is all about end-to-end, and most people would want to buy the best integrated system (given equal price performance).
No one company can make a completely interlinked national IPv6 economy happen by itself. Only government can, and only if one leader or lead group speaks for the entire federal government and can generate enough carrots (billions in additional IT spending, hundreds of interesting collaborative international projects to woo allies), and sticks (only equipment that met federal specs would be connected to the IPv6 network). At this point, a Sputnik-sense of urgency is necessary. “Sputnik fear” led to the creation of both NASA and DARPA, and DARPA funded and led the original Internet. Someone in the US government should now lead a similar effort to elicit funds, manage the specifications, interact with the coalition, and speak to industry, government, and academia alike so that a sense of national and global urgency is created. Without its allies, the US will not be able to be a leader of the Internet, and flipping of “chips” the size of entire industries will continue at an accelerated pace, until the US coalition goes from leading almost everything worth leading, to leading almost nothing worth leading.