The year to unlock the Internet

It's hard to believe, but it's been three years since I began talking in these pages about how the Internet is changing user demands and carrier strategies to meet those demands.

We've covered a lot of technical, financial and political ground in the past threeyears, and we have seen some incremental progress. But as I think about it at year-end, not much really has changed.

Three years ago, two basic issues were driving the growth of the Internet: security and quality of service (QOS). They haven't changed. These two "must haves" are now mantra in the communications industry.

And yet, for all the progress the industry has made, a broad sector of American business remains hopeful-but skeptical.

The truth about the Internet is this: The Internet in the 1990s is still less an engine of business than it is an engine of the imagination. Think about it this way: For the first time in a long while, the imagination of company execs is outpacing technology's ability to make dreams come true.

Once you focus on this, another amazing truth surfaces. The market for mission-critical Internet solutions is already here-as long as security and QOS technologies are firmly in place.

Sadly, these requirements for business-caliber Internet have not been nailed down. Sure, there have been great strides in encryption, user access and virtual private networks. But many of these are vendor-specific point solutions.

It's not all bad news, though. Happily, many key industrywide initiatives for security and QOS will be coming to the fore in the new year.

IPSec, the next level of Internet security, will be part of the industrywide upgrade to the next version of the worldwide standard Internet protocol. Vendors of firewalls, e-commerce and telecommuter access already are testing IPSec implementations, which means the first wave of commercialization should hit by mid-1998.

Secure electronic transaction is an advanced level of security for e-commerce that marries advanced encryption with user ID technologies. The result is that transactions run over the Internet are protected from hackers, and that users buying or selling on-line can be assured they are doing business with whom they think they are.

Spurred by Visa, MasterCard and American Express, there are some 100 secure electronic transaction pilots ongoing worldwide right now. Certified software for a secure end-to-end Internet purchase should be available by summer 1998.

In addition, a number of promising QOS techniques to let customers tune bandwidth to meet their traffic needs are coming out of R&D on high-capacity router and switching technology. Here's a rundown on those most likely to finally kick in next year.

* Cisco Systems' resource reservation protocol (RSVP) aims to let a router on the Internet "phone ahead" to reserve capacity for important traffic on other routers throughout the Internet path. Cisco already has included RSVP in its latest router software. Look this year for Cisco and third-party software firms to make it easier for existing software to get the performance benefits of RSVP.

* Israel-based start-up IP Highway wants to promote the standardization of new Internet policy control protocols, which enable multivendor and multprovider dynamic QOS usage policies. Policy control protocols would adapt to changing network conditions. For instance, as routers clog, loads get large or pieces of the Internet fail, these policy control protocols would re-route traffic accordingly. Watch for these plans to become formal Internet proposals in 1998.

* Many firms are combining Internet and private network techniques to improve QOS. Honolulu-based Digital Island, for example, has created a worldwide "overnet" by combining the company's own star-topology private network with an Internet applications engine specifically designed for deploying business-critical applications. The result is a network that can deliver traffic worldwide over a single router/switch hop.