A journal of IEEE and CAA , publishes high-quality papers in English on original theoretical/experimental research and development in all areas of automation

【Featured Article】A case for quiet: Scientists develop better NICU noise control system

  • Date:
  • 2019-09-24
  • Visited:
  • 8
  • Share:

Babies need soft, soothing sounds. Harsh, stressful noises can delay or even damage their development—especially in premature or ill infants hospitalized in neonatal intensive care units (NICU). To make hospital stays a little more restful for the most vulnerable patients, a team of researchers from Northern Illinois University have developed a system to better control the noise they hear.

-----------------------------------------------------------------------------------------------------------

Paper Information

Lichuan Liu, Yang Li and Sen M. Kuo, "Feed-Forward Active Noise Control System Using Microphone Array," IEEE/CAA J. Autom. Sinica, vol. 5, no. 5, pp. 946-952, Sep. 2018.

Fulltext:

https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8405351

http://www.ieee-jas.org/en/article/doi/10.1109/JAS.2018.7511171

-----------------------------------------------------------------------------------------------------------

It’s easy to control noise when the source position is known. In noise-canceling headphones, for example, a microphone sits upstream from the speaker over each ear. The microphone listens for ambient sound and then triggers an “anti-noise” to cancel out the ambient noise, so the listener only hears the intended music.

The process gets trickier when the noise source is unknown or when the noise source moves, like the constant motion of a NICU. Paper author Lichuan Liu, an associate professor of electrical engineering and director of the Digital Signaling Processing Laboratory at Northern Illinois University, hypothesized that a tool used in things like speech recognition software could be modified to track and muffle more complicated, moving noises.

Called a microphone array, the system consists of multiple microphones pointing in different directions. Usually, as a person speaks, the noise will hit the closer microphones first, allowing the system to pinpoint the location of the noise source and emit the anti-noise signal in the correct direction.

Unfortunately, on its own, the microphone array system doesn’t work as well with highly uncorrelated sounds, such as the doctor pages, nurse voices, or machine beeps that might infiltrate an incubator, according to Liu.

Even with multiple microphones, the traditional algorithm used to isolate the sound source can’t compete multiple sounds at once. The variety of other sounds is too great to cancel them all out.

“In this paper, we proposed a method to integrate a microphone array technique with the feed-forward active control system… to reduce the unwanted noise level,” Liu said.

Liu and her team placed four microphones around an infant incubator and simulated the noise of a neonatal intensive care unit (NICU) by playing recording over loudspeakers. To identify the source of the noise, two algorithms were used in tandem to crosscheck which microphone was physically closest to the source to be able to respond quickly while also keeping the anti-noise signal moving to follow the source of the noise as it moved.

With the successful completion of computer simulations and the incubator experiment, Liu and her team plan to further develop their noise control system and study how well it works in a real, active NICU.


IEEE/CAA Journal of Automatica Sinica

  • CiteScore 2018: 5.31
    Rank:Top 9% (Category of Control and Systems Engineering), Top 10% (Categories of Information System and Artificial Intelligence)