
Combining multiple measurements to reduce noise and improve the accuracy of a signal.Signal averaging is a technique used to extract a weak, repetitive signal from a noisy background by summing multiple measurements of the same event. The underlying principle is that random noise tends to cancel out when averaged, while the signal, which is consistent across measurements, adds coherently. This method is fundamental in many areas of science and engineering, from detecting gravitational waves to improving the resolution of medical imaging.
Signal averaging has been used since the early days of electronics, but it became a standard technique in the mid-20th century with the development of computers that could sum thousands of traces. It was crucial in early radio astronomy and in the discovery of pulsars in 1967.
The signal-to-noise ratio (SNR) of a single measurement is often poor. By taking N repeated measurements of the same signal and summing them, the signal amplitude grows linearly with N, while the noise, being random, grows only as the square root of N. Thus, the SNR improves as the square root of N. This works only if the signal is time-locked to a trigger and the noise is random and uncorrelated with the signal. In practice, the measurements are aligned in time, summed, and then divided by N to get the average.
Signal averaging is used in electrocardiography (ECG) to detect tiny cardiac signals, in magnetic resonance imaging (MRI) to improve image quality, and in gravitational wave detectors like LIGO to extract the faint signals from black hole mergers.
The first pulsar was discovered because Jocelyn Bell Burnell noticed a signal that appeared every 1.337 seconds, which was only visible after averaging many minutes of data, and it was initially nicknamed 'LGM-1' (Little Green Men) because it was so regular that it was thought to be artificial.














