One of the first NCR prototypes of a wireless LAN card circa 1987.
Word of the project attracted Bruce Tuch, an RF engineer who retains his native Brooklyn accent despite living in the Netherlands since 1981. At the time he was working for Philips in Eindhoven, trying to move to Utrecht where his wife was studying.
“NCR wanted an analog engineer which was not so interesting to me but they said if you take the job you could also work on this new project,” said Tuch who became head of engineering for the effort. “My challenge was to get it to market fast and at low cost,” he said.
Of two techniques approved by the FCC, Tuch’s team concluded the direct sequence spread spectrum approach would yield the highest data rates. Military implementations of the technique were well known but used complex codes that would require expensive processors.
Tuch’s team came up with the smallest code size possible. He flew to Washington D.C., got FCC officials to approve the approach and started work on the first prototype.
Meanwhile, Michael Masleid of Inland Steel worked on modelling how wireless signals behaved indoors, especially in the car factories of his customers where lots of metal surfaces would generate reflections. Little information was available on the topic at the time, so Tuch also commissioned wireless researcher Ted Rappaport then at Virginia Tech to conduct separate studies.
With prototypes in the works, Hayes decided the best way to start a standards effort would be to work with an existing group. In July 1988, Hayes and Tuch went to an IEEE meeting working on a wireless extension to the token bus used in industrial automation, but the chairperson didn’t show up.
“They told me if I wanted to do some work, I needed to take the chair, so that’s what I did,” Hayes said.
After about two years’ work, it was becoming clear the token-bus network was not well suited to a gigahertz-class wireless implementation. “One night in our hotel room, I told Vic this protocol was going nowhere, but he knew how to coral the standards bureaucracy, so I told him he should start a new group,” recalled Tuch.
Hayes again rose to the challenge and in September 1990 convened the first IEEE 802.11 meeting. Initially few saw it’s potential. “First, I went to the Ethernet group, but they were not interested in wireless extension,” he said.
Next page: A product is born
A second NCR prototype designed in 1988.