|
|
|
|
| The EU Expert Conference on IPv6
in Vienna, 2006
The recent European Union (EU) Expert Conference on IPv6, titled, Convergence: New Opportunities for Accelerating the IPv6 Momentum, was held in Vienna, at the headquarters of Telekom Austria, one of the stalwart supporters of v6 in Europe. The event was well organized, had some really great speakers, and was an eye opener in more ways than one. The following is a sample of some experiences from the conference. Ulf Dahlsten, a Director of the European Commission (EC), started the conference with a quick overview, noting that the uptake of IPv6 in Europe has been slower than in other parts of the world (i.e., Asia). He noted that the EC has supported 14 large-scale research projects in the past few years (for over 170 million euros), and that IPv6 at the network level is now a mature technology. Reading between the lines, it would appear that the Commission is satisfied that it has completed its support of the R&D sector for IPv6, and would now like to see the technology proceed beyond primarily academic networks into the European mainstream. One could imply (note the word “accelerating” in the event title) that the EC would like to see this transition speed up somewhat. Mr. Dahlsten’s thoughts were echoed by Rudolf Strohmeier, the Head of the Reding Cabinet of the EC. He noted that the EC had conducted a study of both the drivers (he called them “triggers”) and blocking factors affecting v6 in Europe. The three most desirable drivers were: transparency, convergence and mobility (that can be enabled by IPv6). He mentioned that IPv6 would restore E2E (End to End) connectivity, a vital feature of the original Internet. The opportunity of v6 was thus clearly recognized by the IT leaders that were surveyed. However, the study showed that three blocking factors form a chasm that must be crossed first, a vicious circle of: Lack of Demand, Lack of Services and Lack of Applications. These feed on each other, so that no applications results in no demand, which in turn leads to ISPs not providing v6 services, which leads to no applications, et cetera. Mr. Strohmeier added that survey respondents felt that there could be tremendous growth in four to five years, but that only new applications afforded by IPv6 would “make or break” widespread adoption in Europe. Mr. Strohmeier mentioned that Europe needs a new impetus, such as a stimulation of public demand. He noted that the most proactive stakeholders may not be the traditional Internet movers and shakers, but instead could be relatively new players such as the European Ministries of Defense, the transportation sector (including automotive telematics), and the mobile industry, all of which need “future proof” features urgently. The speakers were not limited to European countries. Dr. Hiroshi Esaki, the Executive Director of the IPv6 Promotion Council of Japan, gave a thorough overview of the IPv6 market in Japan. He stated that Japan’s products industry had found 5 really important contributions from IPv6 that may not be fully appreciated elsewhere:
Dr. Esaki stated that the KAME project was now complete, and the resources for that project are being reallocated to mobile IPv6 and other programs. KAME was the BSD-based project, while its brother project USAGI was for Linux. He listed a number of IPv6 applications, including InternetCar and InternetTrain, a building with IPv6-based energy management (there are over 15 of these buildings now in Japan), office machines, and consumer electronics such as television sets; some of these products, such as web vidcams and office printers by Panasonic, are already in their 2nd or 3rd generation. He mentioned that buildings that control their HVAC (Heating, Ventilation and Air Conditioning) with IPv6-enabled sensors have shown savings of 30-40%, meaning that, “With the savings, you could tear down the building and rebuild it every 20 years!” One of the more successful applications is SIP(Session Initiation Protocol)-based VoIP with Freebit, which now has 280 sites at universities, providing PSTN-like quality and reduced calling costs to thousands of dormitory rooms. Dr. Esaki noted that this was also a good example of how IPv6 can simplify design and redesign – the entire Freebit system only required three installation manuals. One interesting revelation was how time horizons can affect a company’s decision to get into IPv6. For instance, Dr. Esaki mentioned that the technology managers of KDDI, the second largest telco in Japan, actually hated going over to IP-based communications; they preferred TDM (Time Division Multiplex). However, because their business planning horizon is 10 years, they conceded that there was no alternative to IPv6, because TDM switches may not be available in 10 years. In contrast, other companies, such as some of those belonging to the European Broadcasting Union (EBU), have business planning horizons of only two years, and came to the conclusion that IPv4 would still fulfill their requirements within that time period. One could speculate that similar short-term thinking may be hampering many American companies. Dr. Esaki discussed emergency response management, and how protection against and recovery from disaster is becoming a very important business area in Japan. He said that sensornets that detect rain and flooding are being installed in various Asian countries, and noted that installing and monitoring these is a “good project for local high school and university students.” He also noted in passing that although less than 5% of Japanese Internet traffic is known to be IPv6, this does not include NTT West, which carries half the traffic in Japan, but does not release figures on IPv6. He speculated that if these figures were known, the total percentage of Japanese traffic that is IPv6 would be a great deal higher. David Wood of the EBU, Ralf Schaefer of Thomson and Dr. Ulrich Reimers of Braunschweig Technical University discussed IPTV (Television over Internet Protocol) in Europe. The bottom line: IPv6 is not yet an issue for the broadcast industry in Europe, and in-home STBs (Set-Top Boxes) are expected to support only IPv4 for the foreseeable future. Mr. Wood noted that the increase in demand for streaming video from VLogs and Podcasts may form a future “rupture scenario.” At present Europe has twice as many Internet-based radio stations as over-the-air stations. Streaming video can prove extremely expensive right now, as viewers tend to watch (or at least leave their display turned on) for extended periods, and only a few ISPs are supporting IPv4-based multicast. They are experimenting with Peer-to-Peer (P2P), which has its own problems but can bring costs way down. Mr. Wood mentioned that DRM (Digital Rights Management) is a major problem in Europe, as a station may have the right to broadcast content only within a certain part of Europe (and would be forbidden from allowing access to the US of such content). The lookup tables for the details of such DRM need to be updated frequently, he noted, and this type of geolocation tracing of Internet-based content presents a “gigantic problem” for broadcasters in Europe. He believes that IPTV will grow, pointing to popular content such as Rocketboom and iTunes (he noted that the iPod was invented by a European engineer, who took it to Phillips, which evaluated it as worthless – his next stop was Apple). Mr. Wood believed that IPv6 could help all these problems with IPTV, but posed the question, “Who will lead Europe forward?” Mr. Shaefer noted that IPTV is well developed in Europe, with over 20 companies participating in it, although these are primarily “walled gardens,” operator-controlled networks within a limited geographical area of subscribers. He said that one thing missing is a “TV Guide” for IP-based content; no one knows how an HNED (Home Network End Device) can easily scan a network to get content updates. He remarked that whereas IPv6 may seem far away for most broadcasters in Europe right now, the market can be fickle; for a long time, MPEG-2 was “good enough” as a video codec, but as soon as the first broadcaster starting using the newer MPEG-4, “everyone had to have it.” Latif Ladid, President of the IPv6 Forum, asked why the EBU opted for IPv4. The answer was, “The European members actually preferred IPv6, but non-European members such as those from the US demanded IPv4.” “However,” Mr. Shaefer commented, “of the 12 handsets in Europe that can show television, nine actually have an IPv6 stack built in for future upgrade capability.” Patrick Grossetete, Cisco’s Program Manager for IPv6, spoke of the different viewpoints held by a network manager, with his infrastructure focus, and an end user, with his application focus. The one must concentrate on enabling his backbone with IPv6, whereas the other is IP agnostic. Mr. Grossetete felt that a promising new applications area would be schools (especially universities), which could use IPv6-enabled features such as content distribution with multicast, class-to-class collaboration, and video surveillance. He mentioned that marketing managers may want to focus on specific customer areas, termed Communities of Interest, that may have common addresses and shared network resources, rather than going for a large but amorphous global market. He noted that Global Addressing does not necessarily involve Global Reachability – that a better business model might involve a community or group with a strong common interest that could leverage the capabilities of IPv6. The Microsoft presentation, by Jonas Persson, the CTO of Microsoft Western Europe, was a real highlight. He touched on various new services that Microsoft is targeting, such as moving the home PC out of the den and into the living room, and seamless networking to integrate service providers and application vendors. He mentioned that the seamless network demands a new paradigm, with ease of connectivity, security and mobility, and that IPv6 is required to support these. The audience listened raptly to the description of Vista, with significant components that are only available with IPv6, and that IPv6 is on by default in the OS, and is preferred. Microsoft has been running IPv6 on its campus for a while, and has learned some lessons. These include: moving applications is the first step, a phased approach is necessary, and transition technologies such as ISATAP can work very well and be cost effective. Professor Jianping Wu, the Director of the China Education and Research Network (CERNET), gave a presentation that was interesting just in its pure scope – CERNET involves 1,500 universities and institutes, 200 cities and over 20 million users, and is the largest native v6 network in the world. He reviewed examples of how strongly China is involved in the Internet, including the fact that 80% of routers are Chinese. Although China appears to be totally committed to IPv6, Dr. Wu still had question marks on some of his slides – the biggest one being “Money???” (i.e., what are the profitable business models and applications for IPv6?). China has a new transition strategy, to get into a native v6 network with v4 going over v6, rather than a dual-stack approach. “We believe this is the best transition strategy, to move 4 onto 6,” noted Dr. Wu. Because this causes problems in identifying the source of v4-based messages, China has asked the IETF to form a new working group to solve this problem. “We must know where every packet came from. Other strategies focus only on ‘Where is it going?’, but that breaks security,” he said. China is investigating applications such as P2P models, IPTV, and WLAN communications; last year, China supported some 70 products, including WLAN terminals, cell phones and medical products. Asked about the tie-in between v6 and 3G, Dr. Wu smiled and said, “I don’t believe in 3G. Sorry.” Rob Goode, Principal Scientist of the NATO Consultation, Command and Control Agency (NC3A), gave an interesting overview of the Alliance’s interest in IPv6, including the desire of moving from multiple special networks to Intranet encryption at the IP layers (with voice and data via IP), as well as reaching Network Enabled Capability, a Service-Oriented Architecture (SOA) approach over a ubiquitous unclassified IP network throughout the NATO area. Desirable new capabilities via v6 included cryptographically generated addresses, auto-configuration and better mobile communications. France and Germany are among the members that have directed future IT equipment should be dual-capable. The Alliance has chosen to go dual stack, with v6 tunneling through v4 until v6 becomes more widely available. NATO is evolving its transition plan, and will coordinate it via a new IPv6 Transition Office. One of the major initiatives dealing with the security aspects of IPv6 is an eight nations project called Interoperable Networks for Secure Communications (http://insc.nodeca.mil.no). Emmanouel Varvarigos, the Director of Networking Technology for the Greek Computer Technology Institute, covered an interesting project to introduce IPv6 to 10,000 schools in Greece. At present each school has both a NAT and a PAT (Private Address Translator), and it is very difficult to debug interconnection problems. Greece is looking forward to exposing its students to the new IPv6 technology. “After all,” noted Mr. Varvarigos, “these are our future engineers.” At present, five of the eight POPs in the Greek school network are dual stack already, and the entire network is being slowly upgraded. They have started to enable IPv6 in a test bed of 50 schools, and will then expand to the 3,000 schools that have broadband. He strongly believes in pilot projects, to gain experience and decide on deployment strategy. They have learned that it really helps in transitioning to v6 if it is part of a general equipment upgrade initiative, and that IPv6 deployment costs can actually be quite reasonable. Eivan Cerasi of Eurocontrol, an organization that manages Air Traffic Control in much of Central Europe, reported that in the future aircraft in that sector of air space will be controlled by VoIP via IPv6. He said that their tests showed that they could not handle NATs for their environment, which involves low data rates but has very high security requirements, and thus see no alternative to IPv6 for VoIP. Air Traffic Control at present consists largely of analog voice services. This will change to voice and eventually data via IP. They are presently migrating from X.25 to IP, and will have mixed X.25/IP for a while. An upcoming standard, the Flight Message Transfer Protocol, will mandate IPv6. Armin Sumesgutner, Head of Strategic Portfolio and Innovation Management, and Dr. Wolfgang Fleischer, Department Head, Packet Switched Core, covered the road map of Telekom Austria. It was heartening to see that this relatively small country appears to have a clear vision of the future, and is emphatically setting up an infrastructure for v6. Another (non-European) country that similarly has a very clear road map for v6 infrastructure is South Korea, which was represented by an associate of Kwanbok Jo, Deputy Director of Infrastructure Policy of the Ministry of Communication. Korea is in many ways the most Internet-savvy country on the planet, with over 72% of homes having broadband connection, and a plan to have a dominant countrywide IPv6 network by 2010. What both of these countries also share is a lack of popular IPv6 applications – both speakers mentioned that ISPs and vendors are still reluctant to invest in the technology, given what they see as market uncertainty. One presenter that was clear on applications was Gérard Segarra, Head of the Information Systems and Telematic Research and Innovative Projects Unit of Renault (which also owns Nissan of Japan). When this automotive group speaks about putting telematics into cars it is with some authority, as they make six million new vehicles each year, and have some 50 million on the road right now. Mr. Segarra outlined the relevant IPv6 advantages for vehicles, which included lots of addresses (each car may need hundreds of them), IPSec, and mobility –– and high-speed mobility, at that. He said that his company is part of a consortium of 50 that includes other large auto makers such as BMW, and has a 22 million euro budget (11 million from the EC) and seven test sites. Telematics can take place between an organization and the vehicle, between the parts of the vehicle, between two or more vehicles (V2V could communicate on traffic jams ahead, for instance), between roadside stations and the vehicle (such stations could establish “virtual traffic cones”), and between roadside stations (to warn of traffic dangers or weather conditions ahead). Applications include monitoring a vehicle’s health, reporting accidents (including the severity of the accident and its location), and providing entertainment content and communications for the driver. Mr. Segarra noted that one advantage to this project is that it is all “Green Field” – it is a brand new capability, and is thus a much easier opportunity to introduce IPv6 rather than an environment that already has legacy applications running in it. Projected near term costs for a suite of IPv6 telematics in a car is about 300 euros, but this should come down in the future, with economies of scale reducing the price. One problem is that some geographical areas only have GPRS offered by the telcos, which could lead to reduced features versus true broadband WLAN connectivity, and the automotive industry is therefore trying to get better service from the telephone carriers – it is interesting to note that the automotive industry may become one of the “non-traditional” prime movers for IPv6, just as Ulf Dahlsten and Rudolf Strohmeier predicted. In summary, this was a valuable and well-produced conference. A great deal of work has been done in Europe (and in Asia) in testing and creating effective IPv6 infrastructures. What appears to be badly needed now is business models and proven ways of making money from new and distinct (i.e., “cannot be done as well with IPv4”) applications. There appears to be some hesitancy among companies in the market to be the first to jump in and invest in major IPv6 consumer applications – several presenters used the term, “People are sitting around waiting for the others to move.” There was also some hope that the US would start to forge ahead – the OMB and Department of Defense directives were mentioned by several speakers. On the other hand, there are clearly very strong and purposeful initiatives in Europe to apply IPv6 in major applications, such as for Air Traffic Control, automotive vehicle telematics and in communication and collaboration projects for schools. What also appears to be needed is some type of demonstration facility
or “Show and Tell” center – outside of IPv6 conferences,
there is no place you can presently go to in Europe in order to see profitable
and desirable IPv6 applications at work, in order to convince political
leaders, the Press and the general public of the feasibility and profit
inherent in the features IPv6 can provide. A demonstration center of this
type is under development in the US, as part of the “IPv6 City of
the Future” project; perhaps a “Sister IPv6 City of the Future”
in Europe might not be a bad idea. |
|
|
