Humanity has been trying to contact extraterrestrial civilizations for nearly 50 years. Even longer than she listens to space in the hope of picking up a message from elsewhere. But it is very difficult to hear something of which we are unaware even of its form. Today, researchers are working on algorithms that could advance our knowledge in this area.
You will also be interested
[EN VIDÉO] How extraterrestrials discovered our existence Researchers from Cornell University (USA) and the American Museum of Natural History claim that many extraterrestrial civilizations – if they exist – have already discovered the existence of our Earth. By simply applying the transit method. As we do ourselves to flush out exoplanets. (in English) © Cornell University
At the very beginning of the 19thand century, the physicist German Carl Friedrich Gauss had suggested that we could be spotted by intelligent extraterrestrial civilizations by tracing reinforcement of corn, mathematical patterns in a region cleared for this purpose in the Siberian forest. A little later, Percival Lowell, theastronomer American amateur, suggested setting fire to oil-filled canals in the Sahara desert. Today the researchers wonder what ideas the extraterrestrials might have had, on their side. To make themselves visible to our eyes of humans.
An international team led by the University of California at Berkeley and the Seti Institute (United States) has developed a new machine learning tool intended to simulate what a message from a intelligence extraterrestrial. A tool called Setigen and which presents itself as a library open-source likely to be a game-changer for future searches for extraterrestrial intelligences.
Let us recall that in the matterEarth scientists first looked for signals radio whose origin is clearly artificial. It was the objective of ozma project directed in 1960 by the famous Cornell Frank Drake. Research then focused on the frequency at which hydrogen emits radiation into interstellar space: 1.42 gigahertz (GHz). A widespread natural broadcast, assumed to be known by any potential intelligent extraterrestrial civilization who would use it to maximize the chances of its message being heard. But nothing.
It’s hard to imagine a technological activity different from ours
So the search extended. To other frequency ranges — over a multi-gigahertz band — to different regions of the sky, and even to other types of signals. And still nothing. All of this comes up against a major obstacle: the fact that humanity has never yet recorded an extraterrestrial signal. A signal on which we could base our experiment. And that would better guide researchers.
Algorithms based onmachine learning have begun to be used to distinguish possible extraterrestrial transmissions from background noise. The best known serves the program BreakthroughListen. It aims to identify continuous signals such as those that are easy and inexpensive in energy to produce on Earth. So always looking for signs of technological activity as we know it.
It is to overcome this difficulty that researchers have now developed a library intended to be regularly updated according in particular to the advances made possible by the integration in the search for intelligent extraterrestrial civilizations of increasingly more perfected. This library should facilitate the production of synthetic extraterrestrial signals. These can then be used as such or added to observation data. Enough to produce large data sets of synthetic signals to analyze the sensitivity of existing algorithms or to serve as a basis for machine learning. And develop new research methods. The ultimate goal is to successfully establish parameters that can overcome the assumptions made by researchers about what an extraterrestrial message might look like.
Interested in what you just read?