A Unifying Approach for Disturbance Cancellation and Target Detection in Passive Radar Using OFDM

This paper addresses the problem of mobile target detection in multipath scenarios with a passive radar using DVB-T transmitters of opportunity. For such emissions, it has been shown the interest in implementing ``mismatched'' correlators, reducing both the zero Doppler contribution (ZDC) masking effects and the false alarm rate. A very efficient mismatched reference signal is obtained with the reciprocal filter (or inverse filter) which consists in a modulus frequential equalization of the transmitted signal. We propose here to revisit the reciprocal filter-based correlator and to reinterpret it as a so-called Doppler channel detector (CHAD). This new interpretation allows a direct rejection of the ZDC, unifying in one and the same step the main disturbance mitigation and the detector construction. We provide a statistical theoretical study of the performance and a comparison with the matched correlator, i.e., the classical cross-ambiguity function (CAF). We demonstrate that CHAD has a random pedestal (a clutter floor level) significantly lower than that of the classical CAF for low Doppler frequency shifts. Numerical experiments on simulated and real data as well validate the mathematical derivations.