Statistical modeling of mains zero crossing variation in powerline communication

In this paper, authors investigate statistical modeling of measured mains Zero Crossing (ZC) non idealities in narrowband powerline communication systems (PLC) using time series approach based on autoregressive moving average model (ARMA). For this purpose, Box Jenkins analysis steps are deployed for detected ZC impulsive variations. Estimated models exhibit a good fit based on R-square metric reaching 90%. This modeling approach is extended to three measurement environment: in home, in lab and rural locations. Obtained models are useful for NB-PLC channel emulator implementation.