The Net Gets a New Look: Japanese

Back to Contents of Issue: October 2000

by Daniel Scuka

THE INTERNET AS WE know it is a creaking edifice put together some 30 years ago by early researchers as a means for a few highly specialized military, and later academic, computers to share data. Back then, packet encryption, bulletproof security, addressing for millions of Web sites, and video streaming weren't even minor considerations. And ecommerce? It hadn't been invented yet. In other words, it's time to trade in the old Internet.

IPv6 (Internet Protocol, version 6), an umbrella of new standards and protocols, promises to give the old Net a massive face lift, allowing for everything from a vastly expanded address pool (over 1 billion) to Net users being able to have their data delivered faster than anyone else's.

Japanese researchers and industry are playing a central role in IPv6 development. Ground zero is the WIDE (Widely Integrated Distributed Environments) program, based at Keio University and led by Internet maverick Professor Jun Murai. WIDE oversees a number of Net-related research projects (see the WIDE sidebar in "University-Industry Cooperation," page 26, July 2000), one of which is KAME, a full-blown effort to produce the textbook standard networking software that will allow IPv6 to be implemented on Unix computers. Many WIDE members here and abroad use the KAME networking stack to connect to the 6Bone network, an experimental implementation of a true IPv6 version of the Internet. "WIDE is trying to implement its own version of IPv6, called KAME, and this code is stable and well regarded," says Mitsuru Maekawa, marketing manager at Nortel Networks Japan. He predicts that KAME's solution will eventually be adopted by the Internet Engineering Task Force (IETF) as the global standard: "IPv6 is defined by the IETF. The KAME code has almost proven global interoperability."

The heart of the IP remake lies in the expanded addressing that will be possible under the new protocol. The current standard (IPv4 -- no one's too sure why IPv5 was skipped), provides for numerical IP addresses that are four bytes long. IPv6 will boost that to 16 bytes, and allow for additional information to be transmitted with each data packet, including encryption and other information that will boost security and quality of service (QoS) significantly. To improve QoS, certain packets will be treated more equally than others. Real-time data, such as from voice or video conferencing, will speed along at top speed, while lower-priority but still mission-critical stuff -- which users will likely be willing to pay a premium for -- will move along at the next priority. Plain old email and file transfers will come dead last, but even those could be sped up if a sender is willing to bear the cost.

And this time, Japan isn't lagging behind in the rush to formalize and help develop the new standards, network protocols, or hardware and software specifications that will be required.

"One reason why many Japanese Internet people are supporting IPv6 research is that there are very few IP addresses left under the current IPv4 standard," says Maekawa, "and Japan, like China and other countries, is suffering from the shortfall."

Maekawa is referring to the fact that, although it was by no means a planned act of electronic imperialism, in the early days, IP addresses were parceled out in large, inefficient blocks, primarily to the US and Europe. "People in Japan think that the Internet was initiated by Americans. Japanese want the Internet to become more international, and Japan wants to be one of the new initiators," says Maekawa.

Note: The function "email this page" is currently not supported for this page.