Engineers were puzzled by the problem of noise in the lines. The current carries the information in “discrete packets” in the lines. But some spontaneous noise could not be eliminated. Sometimes it would wipe out the signal, creating an error. The interference was random, yet it occurred in clusters. I started to investigate by working with time, dividing the day into two, then dividing that period into two, and so on. He found an hour with no errors. But when he divided the hour with errors into two, he found again, a period with no errors and a period with errors. Again, when he divided the
...more
