Study of Weak Signal Detection in Chaos Based on Volterra Filter

Based on the reconstruction of the chaotic dynamic space and the updating capability of adaptive techniques with nonlinearity of Volterra series. A third-order Volterra filter which is used to detect weak target signal in chaos is investigated. And its SVDPARAFAC approach is proposed. The method of using the singular value decomposition (SVD) and parallel factor (PARAFAC) decomposition to factor second and third order kernels is introduced in detail. The experimental results show this method has much better prediction performance for chaotic flow than nonlinear normalized least mean square (NLMS) adaptive Volterra filter and can detect out a very weak target signal in chaos when SNR gets to-70 dB.