Drone Detection Using 4G-LTE-Based Passive Radar

This article presents results on drone detection in a passive radar framework, which exploits 4G long-term evolution signals. The results are based on an experiment using a noncooperative transmitter. The proposed processing was developed from a general passive radar algorithm, modified to account for both the specificities of long-term evolution signals and experimental observations. In particular, it is shown that the fluctuations of the ambiguity function can be mitigated by using the symbols containing pilots for range–Doppler, thus guaranteeing a more stable time– frequency content. This results in improved detection performance. An analysis of the eigenvalues of the covariance matrix, used for an angular–frequency clutter rejection, exhibits the presence of several distinct superimposed signals, complexifying the processing. In particular, it complicates the decoding of the reference signal, limiting the choices of the clutter rejection method. To address the rejection, the choice of an angular–frequency method rather than a temporal-based one (such as extended cancellation algorithm) is then discussed. Both approaches are applied to real data for comparison. Detection results are finally analyzed in various configurations, providing insights on experimental limitations.