High Frequency EMI Data Set


Our experiments for EMI measurements were done at two different test locations: one in a controlled lab settings and another in a residential apartment in Delhi (India).


This work highlights an extensive empirical study of conducted EMI, performed on a set of 24 loads with 4 different test setups in lab settings and with one test setup in home settings.

HFED is a high frequency EMI dataset having traces taken from signal analyser and USRP.

Our data processing and visualization script is open source and is accessible on Github.


Our data was collected from two different locations. One from lab settings using signal analyser and USRP and second set from one residential apartment using USRP. Spice models used to model appliances can be download from the link below. Also a readme is shared in this folder, having references for the design equations followed for these spice models.

NB: Please note link for second set of data from residential apartment is inactive right now as the volume of data is more and it requires sanitization but I will update it soon [Last Update: 09-02-2016, added spice models].


HFED was released as a part of the following paper:

An In Depth Study into Using EMI Signatures for Appliance Identification

Authors: Manoj Gulati, Shobha Sundar Ram, Amarjeet Singh


title={An In Depth Study into Using EMI Signatures for Appliance Identification},
author={Gulati, Manoj and Sundar Ram, Shobha and Singh, Amarjeet},
booktitle={Proceedings of the First ACM International Conference on Embedded Systems For Energy-Efficient Buildings},
organization={ACM} }