How can Noise Make Signals Detectable?

17-09-2020 |   |  By Robin Mitchell

When it comes to signal processing, noise is something that engineers go out of their way to reduce as much as possible. However, adding noise can sometimes be beneficial and allow weak signals to be more easily detectable. What is Stochastic Resonance, how does it allow easy detection of weak signals, and how can it make better sensors?

Why engineers reduce noise?

If there is one thing that engineers hate most, its noise. Noise is the naturally occurring randomness in all electrical and radio systems, and it cannot be removed entirely. Noise in circuits comes in different forms, whether it is due to the randomness of electrons, or due to external radio sources that induce erratic currents. Either way, the nature of the noise is random, and thus it cannot be magically removed from a signal. 

Noise is an issue for engineers as it superimposes itself to signals, and this makes it harder to detect. An excellent example of how noise can interfere with signals is a noisy restaurant; if all customers are talking loudly, it can be hard to hear others in general conversation. Just like electrical noise, you can only hear others in a noisy place by speaking louder. Thus, noise in electrical systems requires signals to be larger by comparison to the background noise for them to be easily detected.

What is Stochastic Resonance?

Stochastic resonance is the phenomena in which a signal that is often too weak to be detected is amplified by noise and thus becomes detectable. Unlike simple addition of signals, Stochastic Resonance relies on the weak signal resonating with noise which results in the original signal being amplified. However, the addition of the weak signal to the background noise also raises the overall signal levels to those greater than a minimum, which the original signal falls below, thus initially making it undetectable.


How is Stochastic Resonance used in nature?

While Stochastic Resonance has not been shown to be utilised intentionally, scientists have shown that some animals can take advantage of its effects. The first instance of Stochastic Resonance being utilised is the Jewel Beetle. In a paper released in 2015, scientists discovered that Jewel Beetles, which rely on forest fires to lay eggs in burned wood, could detect fires much further when a smaller, non-forest fire, sat between the beetle and the forest fire. Typically, the Jewel Beetle can detect forest fires 12km, and the beetles are only attracted to forest fires. However, a small oil tank fire between the beetle and a burning forest allows for detection up to 140km away. It is believed that the oil fire, which the beetles are not attracted to, initiates the IR sensors in the insects, and this allows for very faint signals to be then detected. 

Scientists have also demonstrated that paddlefish, which eat phytoplankton, can have their detection range extended using Stochastic Resonance. Since paddlefish live in muddy waters, they cannot use sight to search for phytoplankton, and thus use electroreceptors. The typical range of these sensors is up to 50 meters, but when electrical noise is induced into the environment, the paddlefish can find phytoplankton up to 100 meters away.

Could adding noise make better sensors?

The effects of Stochastic Resonance in nature make it clear that weak signals can be amplified and shifted to detectable ranges, but could it be used in modern electronics? While it may not help high-speed buses such as PCIe and USB, it could be highly beneficial to sensor technologies. 

This is what a team of researchers hopes with their demonstration of a Molybdenum Disulphide light sensor. Under normal conditions, a light source too dim to be detected cannot be measured by the sensor, but the addition of noise in the sensor allowed for the detection of the light source. It is hoped that such a mechanism can be exploited to improve sensor technologies such as those found in IoT and military, and could even allow for light communication underwater. According to one of the researchers, army personnel has to carry communication equipment that requires large amounts of power to create a reliable link. However, a light-based communication system that can use Stochastic Resonance could be significantly lighter, thus enabling the carrying of more essential supplies. Such technology could also help to create highly sensitive sensors such as environmental monitoring.  

It appears that noise may not be as bad as we all thought, and adding just a bit of noise to a signal could help make very weak signals detectable!

Read More


Profile.jpg

By Robin Mitchell

Robin Mitchell is an electronic engineer who has been involved in electronics since the age of 13. After completing a BEng at the University of Warwick, Robin moved into the field of online content creation developing articles, news pieces, and projects aimed at professionals and makers alike. Currently, Robin runs a small electronics business, MitchElectronics, which produces educational kits and resources.

Related articles