Modeling Research Of Noise In Power Line Based On Wavelet-Markov Chain

The power line communication technology with its unique advantages has become the most widely used means in the world. And it has a good prospect and a broad market apace. However, the low-voltage power line is not the exclusive medium of communication. The noise in the channel of the power line can seriously affect the quality of the communication. In the absence of the relevant standards and reliable analog noise sources, it is difficult to do anti-noise tests. So it is necessary to do research of the model of the power line communication channel noise, which has a important theoretical and engineering significance.The traditional technology of noise modeling is divided into two parts. One is background noise, and the other is impulse noise. The traditional model of background noise can only draw the frequency-domain waveform and the traditional model of impulse noise can only draw the time-domain waveform. Based on the above issues, a new model based on Wavelet-Markov chain is proposed. First, the measured noise is decomposed by wavelet packet transform. Secondly, the peak-type Markov chain model of the wavelet packet coefficients by decomposition is established. The model can construct a new group of the wavelet packet coefficients. Finally, a new set of noise is reconstructed by the new wavelet packet coefficients. It draws not only the time-domain waveform, but also the frequency-domain waveform. Based on two different environments, which one is in the laboratory, and the other is in the catenary of metro, the circuits of measuring noise are built according to the actual situations. By the circuits, the actual noise data is measured. And this measured noise is used to verify the effectiveness of the new model. Based on Wavelet-Markov chain, the language of verilog is used to design modules and to achieve the simulation of the new model. So the preliminary design of the noise generator which provides a theoretical basis is completed.