Performance of multichannel parametric detectors with MCARM data

We consider the problem of detecting a multichannel signal in the presence of spatially and temporally colored disturbances. The parametric Rao and GLRT detectors, recently developed by exploiting a multichannel autoregressive (AR) model for the disturbance, have been shown to perform well with limited or even no range training data. In this paper, we evaluate the detection performance of the parametric Rao and GLRT detectors using airborne radar data obtained from the Multi-Channel Airborne Radar Measurement (MCARM) database. The data contain typical clutter found in airborne radar systems, and cover a variety of scenarios including dense-target or heterogeneous environment. Experimental results with the MCARM data show that the parametric Rao and GLRT detectors can provide good detection performance with limited or even no range training data in real radar environments. As such, these detectors offer useful solutions to detection problems in dense-target or heterogeneous environments.