Radar Coincidence Imaging with Stochastic Frequency Modulated Array

In radar sensing and imaging, the azimuth resolution is a main concern, which is limited by the antenna aperture, and as a result the targets within the beam cannot be distinguished. By enhancing the diversity of radiation, radar can obtain additional information for resolution. In this paper, a high-resolution staring imaging technique named radar coincidence imaging (RCI) is investigated. Originated from the classical optical coincidence imaging, the RCI captures super-resolution in azimuth, which breaks through the Rayleigh resolution limitation of antenna array by modulating the wavefront of transmissions. The spatial resolution of RCI is defined by the spatial correlation function of the stochastic radiation field. A scheme of RCI with a stochastic frequency modulated array using frequency-hopping waveforms is proposed, while the imaging model is established. Three image reconstruction algorithms, i.e. the pseudo-inverse algorithm, Tikhonov regularization method, and sparse reconstruction algorithm, are investigated and compared with respect to targets of different complexity. Performance analysis of these reconstruction methods in the presence of noise is presented by the relative imaging error. Finally, a typical RCI system based on the digital transmitter/receiver array is established. Outfield experiment results verify the effectiveness of the RCI.