Adaptive detection of range-spread targets without secondary data in multichannel autoregressive process

Adaptive detection of range-spread targets without secondary data is addressed in a multichannel autoregressive Gaussian disturbance with unknown space–time covariance matrix, by utilizing the Rao test. The proposed Rao test without secondary data is theoretically proved to be asymptotically (large-sample in the number of temporal observations) constant false alarm rate with respect to unknown space–time covariance matrix, thanks to an asymptotic equivalence between the Rao test and the generalized likelihood ratio test. Moreover, the performance loss due to no secondary data can be remedied by appropriately increasing the temporal dimension. The performance assessment conducted by Monte Carlo simulation, also in comparison with the existing detector without secondary data, confirms the effectiveness of the proposed detectors.