Adaptive Target Detection Techniques for OFDM-Based Passive Radar Exploiting Spatial Diversity

Abstract

Recently, the benefit of using a channel-based detector (CHAD) with a passive radar exploiting orthogonal frequency division multiplexing waveforms radiated, e.g., by DVB-T transmitters of opportunity has been studied. When multiple antennas are available on receive, we state that CHAD can be seen as space–time array-processing performing on a particular coherent frequency datacube. Building on this new interpretation, we propose an improved version of CHAD in the form of fully dimensional space–time adaptive processing (STAP). Optimization of the signal-to-interference-plus-noise ratio is obtained combining a linearly constrained minimum variance space–time adaptive beamforming and a least squares spatial adaptive filtering. Unlike classical STAP approaches, no training data are used here and only one space–time sample matrix inversion is required. The computational load is then highly reduced allowing a practical deployment. Moreover, since no beamscan in space is performed, the knowledge of array characteristics is not required and the performance shall not be impacted by any calibration errors. Finally, all the target returns being located in a single 2D (range-Doppler) surveillance map, the detection process is then simplified. Results on experimental data show the interests of this new surveillance scheme: No need for a prior rejection of the dominant interference, systematic reduction of secondary lobes, and discrimination of slow moving targets.