Appliance Fingerprinting Using Sound from Power Supply

Compared with traditional appliance identification methods, our method uses the acoustic signal that already exists in life, which saves cost and is easy to deploy. Our method can also be employed in many fields like privacy protection, power consumption reducing, and the intelligent enhancement of home appliances. However, there are still challenges to overcome: i) The short detection distance. We discovered that the emission power of some home appliances was often low, resulting in that the ultrasonic signal could only be captured within a short distance. ii) The mixed signal. There are alwaysmultiple household appliances working at the same time, so theultrasonic signal we actually receive is a mixed signal. Even worse,many appliances’ frequency ranges coincide with each other. Inthe future, methods like using high-precision microphones will beemployed to enlarge the detection range. Also, new methods willbe proposed to distinguish various household appliances working simultaneously.