How SETI Works

SETI and You

SETI@home screen (larger version of the image)
SETI@home screen (larger version of the image)

In 1999, University of California at Berkeley scientists Dan Werthimer and David P. Anderson worked on Project SERENDIP. They recognized that a limiting factor in analyzing the data from the Arecibo dish used by SERENDIP was the available computing power. Instead of using one or more large supercomputers to analyze the data, many smaller desktop PCs could be used to analyze small pieces of data over the Internet. They devised a screensaver program called SETI@home that could be downloaded from UC Berkeley over the Internet and reside on a participant's home computer. The program can work in residence or as a screensaver.

Here's how the project works:

  1. Data are collected from the Arecibo dish in Puerto Rico, where Project SERENDIP is presently conducted.
  2. The data are stored on tape or disk along with notes about the observations, such as date, time, sky coordinates and notes about the receiving equipment.
  3. The data are divided into small chunks (approximately 107-second blocks) that desktop PCs can utilize.
  4. The SETI@home program on your PC downloads a chunk data from the computer servers at UC-Berkeley.
  5. Your PC analyzes the chunk of downloaded data according to the algorithms in the SETI@home program. It takes about 10 to 20 hours to analyze the data, depending on the computer's microprocessor and amount of memory.
  6. When finished, your PC uploads its results to the UC-Berkeley servers and flags any possible hits in the analysis.
  7. After the upload, your PC requests another chunk of data from the server, and the process continues.

The screen saver is divided into three sections: the data-analysis window (upper left), the data/user information (upper right) and the frequency-power-time graph of the data as it is being analyzed (bottom). The chunk of data is analyzed by spreading the data out over many channels using a mathematical technique called a Fast Fourier Transform (FFT). If the data are random, then the signal in all of the channels will be equal. If a signal (spike) is present, then one or more FFT channels will stand out above the rest, above a certain power-level threshold. Next, the program looks to see if the frequency of any spike is shifted slightly to other frequencies -- this shift would be caused by the Earth's rotation, indicating that the spike is of extraterrestrial origin. Finally, since the Arecibo dish is stationary -- does not track objects with the Earth's rotation -- an ET signal would drift over the dish's surface, from edge to center to edge, and a plot of the spike over time would look like a bell-shaped curve. The program tests to see if the spike fits this curve. If these three criteria are met, the program flags the information for later analysis by UC-Berkeley.

Data Analysis window of SETI@home Data Analysis window of SETI@home
Data Analysis window of SETI@home

The data/user information section of the screen contains the notes on the observations that obtained the data chunk, as well as notes on the user.

Data/user information portion of the SETI@home screen Data/user information portion of the SETI@home screen
Data/user information portion of the SETI@home screen
Graph window of SETI@home screen Graph window of SETI@home screen
Graph window of SETI@home screen

The graph screen allows the user to see the progress of the analysis at a single glance. The program notes all of the observed spikes and relays this information back to UC Berkeley for further analysis. Each data set is processed independently by two users for corroboration. If a spike passes the criteria for a possible signal, then other SETI projects will examine the coordinates in greater detail to confirm the finding.

With SETI@home, a computer and an Internet connection, you can participate in SETI research. To date, the SETI@home Web site receives one-million hits and 100,000 unique visitors per day.