Modeling Research Of Noise In Low-voltage Power Line Communication Channel

In recent years, the low-voltage power line communication technology has been widely used in various fields. The noise is one of the main factors affecting the reliability of low-voltage power line communication. It is necessary to study the high-precision modeling of low-voltage power line channel noise. The model of low-voltage power line channel noise is studied with the measured noise data in this paper. The results provide a theoretical basis for the establishment of a high-precision model of the low-voltage power line communication channel noise in the future.Characteristics of low-voltage power line channel are described. The low-voltage power line channel noise measurement circuit is built and the actual noise data is measured.The AR model of background noise is established. The parameters of the model are calculated with singular value decomposition, Levinson-Durbin recursive method and Gram-Schmidt orthogonal. Simulation results show that:the results used singular value decomposition is extremely accurate, but the calculation time is very slow; the results used Levinson-Durbin recursive method is very imprecise, but the calculation time is very fast; the results used Gram-Schmidt orthogonal is accurate, and the calculation time is fast.The Markov chain model of impulse noise is established which considering the relationship between the amplitude, interval, duration and symbols. But the model does not accurately reflect the characteristics of amplitude changes. Then the concept of peak-type Markov chain is proposed. The peak-type Markov chain model of impulse noise is established and simulated. Simulation results show that:the two models are all correct, and the peak-type Markov chain model is more accurate than the traditional Markov chain model, it can reflect more characteristics of impulse noise.The general model of the power line noise was established with the AR model of background noise and the peak-type Markov chain model of impulse noise and the model was emulated. The power line noise with periodic pulses was been special handling. The power line noise was divided into the noise within the pulse group and the noise outside the pulse group. The special model of this noise was established and this model was emulated too.