Combined Eigenvector Analysis and Independent Component Analysis For Multi-Component Periodic Interferences Suppression In PRCPM-PD Detection System

Abstract

Interference suppression and information extraction in pseudorandom code phase modulated pulse Doppler detection system is a practical and challenging problem. In order to extract useful range and velocity in the presence of single-channel multi-component periodic interferences, this paper presents an interference suppression method based on eigenvector analysis and independent component analysis (ICA) by leveraging the characteristic of generalized periodicity of received signal. Using the generalized period of the transmitted signal, we divide the observed received signal into different segments and derive the sample covariance matrix. By computing the eigenvalue decomposition of the sample covariance matrix, we obtain all the eigenvectors and find the corresponding eigenvector with the maximum eigenvalue, which is the sum of the basic waveform components with the same generalized period. Then multi-dimensional matrix can be structured by changing the length of observed signal. The multi-dimensional matrix is used to accomplish ICA so as to separate all the components. Finally, the useful echo signal can be reconstructed; the range and velocity information can be obtained. Simulation results show that the proposed method is effective to excise multi-component periodic interferences and obtain the range and velocity information accurately at high signal to interference ratio (SIR). At low SIR, recursive separation is required. Compared with the existing methods, the proposed method has better performance.

Keywords

• Single-channel multi-component periodic interferences suppression
• PRCPM-PD system
• eigenvector analysis
• Independent component analysis