How SETI@Home Screens ET Candidates

SETI@Home is a distributed computing initiative that analyses radio signals for signs of extra terrestrial intelligence. It relies on volunteers who use their own computers to download and crunch data from the Arecibo Radio telescope in Puerto Rico using bespoke software available from the SETI@Home website.

The project’s central challenge is to spot a real ET signal against a background of noise and interference. The software searches for five different types of patterns that are unlikely to be produced by noise, things like three power spikes in a row and pulses that could represent digital signals.

Using these criteria, SETI@Home volunteers have identified some 4.2 billion signals of potential interest. That’s a significant number. But even though these signals have all been screened by the SETI@Home software, most, if not all, of them are likely to be the result of noise or interference.

So today, Eric Korpela at UC Berkeley and a few buddies outline how they analyse these candidates. One important task is to identify common sources of interference that produce intelligent-like signals. “By far, the most common source of interference in the SETI@home data set is radar stations on the island of Puerto Rico,” say Korpela and co.

Most of these and other types of interference can be identified and removed automatically. However the team is hoping to crowdsource the analysis of the signals that slip through using software that trains volunteers to inspect signals visually and identify those that are the result of interference.

One important criteria for ET signals is that they must be persistent in time and frequency. In other words, an interesting signal must be observable in the same area of sky at a later date. To monitor this, Korpela and co and have designed a program called the Near-Time Persistency Checker or NTPCkr that creates a kind of heat map of interesting signals in the sky.

When an interesting candidate turns up, a certain area of sky becomes ‘hot’. If the signal continues, that area of sky remains hot, otherwise it cools down over time. Anything that remains hot over a decent period of time is worth looking at in more detail.

So what has SETI@Home found? Nothing really. The most significant candidate is a sourse called Radio source SHGb02+14a which the team revealed in 2004. But even this is an unconvincing candidate. In the area of sky in which it was found, there are no stars within 1000 light years of Earth and most commentators think the signal is probably due to random variation.

Nevertheless, the search continues.

Ref: arxiv.org/abs/1109.1595: Candidate Identification and Interference Removal in SETI@home