Abstract:

To sdve adaptive detection problem that the cell under test is contaminated by the subspace constrained rank one interference, namely the noise covariance matrix of the test data is superposed by a rank one interference covariance matrix, the interference is assumed to be located in a known subspace, and using the 2-step generalized likelihood ratio test criterion, we derive a subspace constrained rank one interference detector (SC-ROID). We also derive a rank-reduced adaptive coherent estimator (rr-ACE), which exploits the prior information of the interference to reduce the computational load. The computer simulation results show that, under the condition of no interference or no prior information of the interference, the SC-ROID performance amounts to the adaptive coherent estimator detector. In other cases, the SC-ROID outperforms other conventional detectors. The rr-ACE outperforms other competitors when the size of the reference data is low.

Key words: adaptive detection, rank one interference, subspace constrained interference, generalized likelihood ratio test, rank-reduced