Statistical characterization and modelling of impulse noise on indoor narrowband PLC environment

This paper focuses on the characterization and modelling of impulse noise in narrowband power line communication. Two modelling approaches of the measured noise are studied and compared: the memoryless Middleton class-A model and the Markov-Middleton model with memory. Model comparisons with measurements show that Markov-Middleton is compliant with the impulse noise temporal structure since it considers the pulse sample correlation. In the frequency domain, both models do not well represent the frequency behaviour of the impulse noise. Differences in duration and distribution in time of errors generated by measured/modelled noise are noticed.