The PLC Signals' Noise Mitigating Algorithm with PCA

Abstract

Abstract. Noises in the PLCPLC) systems are too severe. In the paper, we proposed a new method to mitigate such noises based on the principal components analyses(PCA). There are 4 steps in this algorithm. Signals generalization is the the first step. the one dimension PLC signal is extended to a multiple dimensions signals. In the second step, the new signals are decorrelated by adding the white noises. For such the uncorrelated data set in the transformed domain, we use the PCA technology to mitigate noises. The useful PLC signals can be recovered with applying the inverse transform matrix as the forth step. We verify this algorithm by the simulations.