SETI scientists do not look for patterns but rather a lack of patterns in a signal. Although this may seem puzzling, its really a matter of physics. The first challenge facing any SETI project is detecting a signal against the background of cosmic and terrestrial noise. A signal containing a great deal of information will be spread across the spectrum more than a very simple signal containing little information would be. An "informative" signal will look more like random noise and thus will be harder to detect. So, in SETI, we look for very simple signals, ones that are easy to distinguish from astrophysical sources.
Natural radio sources like quasars and pulsars are "broadband," meaning they emit over a broad range of frequencies. Communication signals may also be broadband but may contain components that are very narrow and easy to distinguish. For example, an analog TV signal spans about six MHz but contains two carrier signals (video and audio) that are less than one Hz wide. These carrier signals are much narrower than the narrowest known astrophysical source. Therefore, SETI programs searching in the radio spectrum look for very narrow bandwidth signals.
Optical SETI (OSETI) programs must detect a laser signal against the glare of ETs star. The best way for a laser to outshine a star is to do so very quickly, in a very short pulse lasting less than a billionth of a second. OSETI programs thus measure the light from a star in nanosecond intervals looking for a slight, brief excess of photons that could originate from a powerful pulsed laser.
Once a signal is detected, then the hunt for patterns will begin. The detected signal may be a simple beacon that points the way to an information channel. Scientists have speculated that a binary code, a pattern of ones and zeros, will be used, perhaps containing a symbolic language or pictures. In 1974, at the rededication of the Arecibo Observatory, humans sent a binary-coded picture in the direction of the star cluster M13. The radio signal used two radio frequencies to represent the dark and light pixels of the message. Other message-encoding schemes have been suggested, but in truth, scientists really dont know what to expect. That is part of what makes SETI so interesting.
Answer originally published January 12, 2004.