Abstract— Power Line Communication (PLC) is the use of the existing public and private mains power wiring for the transmission of telecommunication signals (data). PLC technology is used for a variety of functions, such as computer networking, utility control systems and smart grid. However, one of the predominant challenges faced by PLC is noise, often provoked by end-user equipment. In this paper, an algorithm to characterize different noise types – peculiar to PLC – is developed. Twelve devices across three load classes (residential, commercial and industrial) are chosen for the study. Experimental measurements are carried out on the devices to capture signals of interest. The characterization process in the work is a combined approach of the Power Spectral Density (PSD) and time stream. PSD analysis is carried out on the captured noise signals using the Welch’s PSD estimation technique. The PSD and other parameters obtained from the time stream are used for the characterization process. Major findings show that most of the devices exhibited either colored background noise, periodic impulsive noise synchronous with the mains or periodic impulsive noise asynchronous with the mains. The outcome of this research will be useful for development and deployment of power line technologies in a manner that mitigates noise problems.