Our Research

FINDING ARTIFICIAL SIGNALS

Lights in the Dark

You’re driving through a desert valley, scanning through the radio dial, but you’re picking up nothing. Just static. The annoying hiss of background noise coming through the speakers. No indication that there’s anyone else around for miles. The road reaches the valley’s edge, and starts to wind its way up a ridge of mountains. You reach the summit, begin your descent, and the radio suddenly crackles into life. At last, a sign of civilization.

What’s really going on here? Somewhere within the detection range of your car radio, a transmitter is encoding information in the form of radio waves. In this case, the encoding scheme is quite simple, and your radio can easily decode the signal to produce the original sound wave. But even in cases where we don’t know what information a signal contains, there are often still indicators as to whether it is artificial.

The above demo shows how information encoded in radio transmissions can be turned not just into sound, but also into a visual representation of the radio spectrum. Here we see how strong the radio signal appears at different frequencies (represented as the line graph towards the top of the screen). The information carried by the signal is encoded in a narrow range (called the bandwidth) of frequencies. In the case of FM transmissions picked up by your car radio, this information represents the audio signal that ends up coming out of your speakers. But even when we tune to frequencies where non-audio data such as pagers or digital data are being broadcast, it’s obvious that these signals are engineered. In part, this is because the signal exhibits patterns that don’t appear natural, but additionally, the bandwidth of artificial signals is often much narrower than that of radio emission from natural processes.

So finding extraterrestrial intelligence ought to be easy. Simply point a sensitive radio detector at the sky, and look for narrow-band signals. Unfortunately, it’s not that easy. Radio frequencies are full of artificial signals generated by humans, many of which look very similar to the kinds of signals SETI experiments are trying to detect, and so radio frequency interference, or RFI, is a major problem for our experiments. Also, we don’t know ahead of time which frequencies ET might be using to transmit. So we have to tune to many “stations” at once.

Frequency Spectrum

This presents a huge signal processing challenge - one that we have been addressing for the last 15 years with the SETI@home software. By distributing chunks of data to millions of volunteers, who run automated signal processing algorithms on their home computers, reporting back promising candidates to be stored in our database, we can search for a large range of possible signals over a wide range of channels.

Until now, data for SETI@home have been coming from just one telescope, the Arecibo Telescope in Puerto Rico. But now, as part of Breakthrough Listen, we are releasing data from the Green Bank Telescope (GBT) in West Virginia to SETI@home users. SETI@home will run as before, but with more data than ever before, as we use GBT simultaneously with ongoing observations, piggy-backing on other astronomy programs as well as observing our own targets.

We’ve also deployed dedicated hardware to GBT as part of Breakthrough Listen - onsite spectrometers that can scan billions of radio channels at once in the search for signals that appear artificial. We’re releasing data from GBT and APF into a public archive, and we're inviting those with sufficient technical experience to download our data, and to come up with creative algorithms and analysis techniques to make the most of this unprecedented dataset.

Help Us Analyze the Data!

Download SETI@home

Read On

Learn about how Breakthrough Listen will search for laser signals using the Automated Planet Finder

Continue to Page 3, APF Info


More information about the kinds of signals SETI@home is searching for:

Detect radio transmissions with your computer