Asynchronous noise removal for earbud-based PPG sensors

This paper addresses motion noise removal for an earbud-based reflective photoplethysmographic (PPG) sensor. This system can be worn comfortably on the ear for a long time. However, body movements such as walking and running cause signal degradation. To improve signal quality, I propose a novel asynchronous noise removal method using only the right and left PPG signals. It cancels not only uncorrelated noise components but also asynchronous noise components by simply introducing a weight function in the calculation process of a Wiener filter. Experimental results demonstrate that heart rate was successfully measured by this system. The simplicity of the procedure and effectiveness of the proposed method may make current systems more reliable.