Detection of Transient Signal Based on AdaptiveSingular Value Decomposition

This paper introduces the principle of?signal detection by Singular Value Decomposition (SVD), and discusses the singular values distribution characteristics of the transient signal and Gauss white noise which received from electromagnetic-pulse along the well bore. Then the definition of singular value increment is given, and the signal is distinguished easily from noise according to those singular value increments differences. On this basis, an adaptive singular value decomposition method based on selection of main singular value using singular value increment is proposed for the detection of transient signal, and the detailed flowchart of the method is illustrated. Simulation results demonstrate that, the proposed method has a good effect on distinguishing the transient faint signal from noise and the performance outperforms the typical SVD.