Power Line Channel Noise Modeling Algorithm Based on IGA Algorithm and BP Neural Network

In order to accurately construct the noise characteristic model of Power Line Communication (PLC) channel, improve the reliability of the communication system, and finally realize safe, fast and accurate communication technology. In this paper, we propose a neural network combined with Immune Genetic Algorithm (IGA) to model and study the channel noise of power lines. In this method, the statistical characteristics of noise are obtained by learning the measured noise samples. Firstly, the actual power line noise sample set was collected in the field and normalized as the input of the neural network. Secondly, the noise features are learned through the continuous training of the back propagation (BP) neural network optimized by the immunogenetic algorithm. Finally, the measured noise data and the data generated by the IGA-BP model are compared and analyzed for each order of statistical characteristics, which verifies the effectiveness of the method.