Dr. Whitfield Diffie, the Chief Security Officer and Distinguished Engineer, Sun Microsystems gave the first keynote address on Tuesday morning. As many of you know, he, along with Martin Hellman, developed a form of cryptography for ordinary computer users. The title of his presentation was "Information Security on its 100th Birthday." His talk was a history of cryptographic techniques over the past 100 years. Radio transmission, begun about 100 years ago, caused a revolution in information security. Before, when messages were carried by hand or at least confined to wires, the need for encryption was limited, although it had been used for hundreds of years. The advent of radio, where an adversary might be receiving the transmission better than the intended recipient, made encryption mandatory for sensitive messages, because that was the only protection. The first attempts to encrypt radio circuits overwhelmed code clerks using codebooks. Improvements were electromechanical in nature into the 1950s. Examples are punched tape, Purple, Enigma and Sigsaly. The Japanese Purple machines used telephone stepping switches. The next wave of improvements, begun in the 1950s, used shift registers. One big problem of secure transmission in those days was key management. The keying material had to be conveyed to all receiving sites by a separate channel, usually in the form of paper. The advance in cryptography introduced by Diffie and Hellman was public key cryptography, which greatly reduced the key management program.
Dr. Diffie gave a brief history of secure voice, from the World War II Sigsaly to the STU III. This spanned a roomful of equipment to secure one voice conversation to the present device not much larger than a telephone. Recent work in cryptography includes multilevel of security in one computer system, implemented by trusted software, that is, software written by cleared software engineers. One of the latest efforts in trusted platform modules is certification of operating system software by attestation, the certification that software functions are not misrepresented to users of a networked computer system.
On Tuesday afternoon, Network World presented its 2003 User Excellence Award to Washington Mutual. Then, John Gallant, President and Editorial Director of Network World hosted a panel discussion having the somewhat puffy title: “Beautiful Minds 2 – Top IT Executives Share Their Insights and Money-Stretching Ideas For Innovating in a Tough Market.” The two panelists were Robert Galey, Chief Information Officer, Amtrak, and David Swartz, Chief Information Officer – Information Systems & Services, George Washington University. Both panelists said budget pressure was still high, but that their budgets were a little up from last year. In 1998, Amtrak invested in IT for bringing on new customers. The 9/11 attacks had the effect of increasing ridership. There is still a tendency in railroad management to adhere to 150-year-old thinking, but Mr. Galey and his staff are working on making every moving train a node on its WAN. Amtrak replaces computers every three years, and email and networking activities are increasing. GWU on the other hand, has experienced shrinkage of its endowment in recent years, because of the stock market decline. Traffic volume requirements double or triple every year. Security is a big part of the job for both panelists. Mr. Galey stated that 85% of the traffic on Amtrak networks is not related to business. He pointed out that it takes about 13 hours for an anti-virus company like McAfee or Norton to react to a new worm or virus, but that traffic volume from a virus can grow to large proportions in a matter of minutes. GWU’s experience is that cookies facilitate break-ins. GWU sequesters computers that have viruses and out-of-date antiviral software. Mr. Galey remarked that Linux would not be the answer to virus vulnerability, because the source code is in the public domain, and hackers will design viruses based on knowledge of this operating system. As to acquisition strategy at Amtrak, it is putting out RFPs for software maintenance, instead of staying with original vendors. Mr. Swartz, on the other hand, is using the economies of large-scale buying through academic consortia. For example, there is the 26-member consortium of colleges and universities in the Washington-Baltimore area sharing a fiber optic ring. GWU is trying to minimize the number of vendors of hardware and software of its systems. Instead of blanketing the entire campus with high-speed Internet access, GWU provides hot spots where students congregate. Amtrak will install Wi-Fi access in six stations in the Northeast Corridor for passenger use this year.
On Wednesday morning, Mr. Howard Anderson, Managing Director, YankeeTek Ventures, gave a presentation titled, "The Next Five Years." He began by noting that Alexander Graham Bell was asked what the telephone might be used for. The answer was that someone would be called to prepare to receive a telegram. The first significant innovation in telephone technology was the voice mailbox, introduced about 100 years after telephone service began. (We are not concerned here with the tremendous technical innovations involved in providing telephone service, such as multiplexing, microwave relays and electronic switching.) In the 1980s, the PBX was going to be the main office console, through which all communication would be controlled. In the 1990s, LANs came in to allow sharing of expensive printers. Now, it’s the open system, to handle voice and data.
Innovations will involve the Internet. Amazon (became profitable this past 4th quarter) is not principally a bookseller. It’s a software factory. It tailors software to allow other retailers to sell via the Internet. Other ventures will follow roughly the same path; we’ll all be "amazoned."
There are four steps to a disruptive technology: denial, anger, grudging acceptance, and capitulation. First, the provider of Internet technology denies to his customers that a certain innovation is what the customer wants, then begins to provide it if the customer insists, then begins to accept the innovation himself, and finally embraces it and recommends it to new customers (capitulation).
In discussing IT budgets, Mr. Anderson said industry insiders knew Y2K was not going to be a problem, but used this scare to get 20% budget increases for IT departments. The money fueled a buying binge.
A longtime feature of COMNET, the Town Meeting, was held Wednesday afternoon. As before, Mr. Richard E. Wiley, former FCC Chairman, hosted a panel discussion of regulatory issues. Of the four panelists, two were legal advisers to FCC commissioners, and the other two were counselors to House and Senate staffs. This is the eighth anniversary of the 1996 Telecommunications Act, and the main topic of discussion was how well it has worked out. The consensus was that the Act does not address the issues of today, that the common carrier of years ago is an outdated concept. The Act was heavily concerned with analog services. Someone noted that the Act was ten years in preparation, and it will take about that many years of groundwork to write a replacement act. One thing agreed upon was that the government should facilitate delivery of services to the consumer. Under Section 271 of the Act, the FCC can give charters to companies to deliver services, and, it was noted, the FCC can retract the Section 271 charters, and will use this as an enforcement tool. Another issue was how much the government will tolerate mergers. If merging begins among the seven Bell companies, the government will probably tolerate three or four companies surviving, but probably not less. Another issue that will require involvement of Congress is VoIP. This has several sub-issues, among them law enforcement access to networks and E911. The government will adopt a hands-off approach to wireless broadband, except in rural areas, a situation reminiscent of the role the Rural Electrification Administration performed in the electric power arena in the 1930s and 1940s.
There was a question-and-answer session at the end, and as often happens, the “questions” are opportunities to make statements. One "questioner" made the point that some countries, South Korea in particular, have more broadband deployment than the U.S. does, because our FCC leadership is mostly lawyers, while the Korean FCC has mostly engineers in command. The FCC panelists’ reply was that there are many engineers in the FCC, and the FCC is trying hard to recruit more engineers. It did not seem to me to be a strong rebuttal to the "question."
Yes, the show was significantly smaller this year, and lightly attended, at least on Tuesday, the first day the exhibition area was open. The cheesecake dolls were fewer, and seemed to be older. The IEEE did not have a booth there this year, and the only other IEEE member in attendance I recognized was Saj Durrani. For the first time, there was no announcement of next year’s conference. That’s too bad, because the show gives access to some of the movers and shakers in the computer networking business, and you see working samples of the new hardware coming on the market. One feature I’d miss is the Reiter’s Scientific & Professional Books display store. Having the latest telecommunication and computer networking books has been a real plus for me, and I plan to buy a book or two every year at COMNET.