Mismatched filter bank design for radio frequency interference suppression in multiple-input multiple-output over-the-horizon radar

Abstract

Radio frequency interference (RFI) is a major problem in Multiple-Input Multiple-Output Over-the-Horizon Radar (MIMOOTHR). Selecting working frequencies and clean bandwidth in the complex electromagnetic environment is challenging using traditional frequency selection methods. We note that MIMOOTHR usually transmits Time-Staggered Linear Frequency Modulated Continuous Wave (TS-LFMCW) waveforms, so the target echo that occurs before matched filtering is a broadband (about tens of KHz) signal, while the RFI after frequency selection is usually a narrowband (about a few KHz) signal. To suppress the narrowband RFI and complete the waveform diversity process, the mismatched filter bank is designed to replace the original matched filter bank without changing the TS-LFMCW waveforms. First, we establish the mismatched filtering optimization problem for MIMOOTHR. We employ the output signal-to-interference-plus-noise ratio maximization criterion of adaptive RFI suppression and two constraints for adjustable target signal attenuation and TS-LFMCW waveforms diversity performance. Then, two analytic solutions to this constrained optimization problem are derived to guarantee the computational efficiency of the mismatched filtering. The proposed mismatched filtering methods can detect the target successfully in the presence of narrowband RFI. The RFI suppression performance, the waveform diversity performance and computational efficiency of the two mismatched filter banks are analysed using numerical results.