A Novel Noise Reduction Method Based on VMD and SVD for Recovered Manchester Coding Signal

Abstract

ABSTRACT Anovel noise reduction method based on variational mode decomposition (VMD) and singular value decomposition (SVD) is proposed to retrieve measurement information from the recovered Manchester coding signals. Firstly, the recovered Manchester coding signals are decomposed into the intrinsic mode functions (IMFs) by VMD. The time-frequency matrix constructed by the IMFs is further decomposed by SVD. Subsequently, the singular values, which are corresponding to the noise component in the signals, are removed with the aid of an appropriate threshold. Finally, the signal is reconstructed from the remained singular values by singular value inverse transform. The feasibility and effectiveness of the novel noise reduction method are verified by simulation signals and real signals. And the complexity of the proposed VMD-SVD method is evaluated. The experiment results show that the proposed method is more effective in the noise reduction for the recovered Manchester coding signals by comparing with the existing methods.