Detection performance degradation due to miscalibrated arrays in airborne radar

Several algorithms have been proposed for the standard radar signal detection problem. Two of these algorithms, the generalized likelihood ratio test (GLRT) and the adaptive matched filter (AMF), are studied in this paper. Approximate distributions for these tests are derived under the signal present hypothesis. The signal model is then generalized to include random array perturbations. The approximate distributions are used to analyze the detection performance of these tests when array perturbations are present. The array perturbations are shown to cause a significant degradation in probability of detection for appreciable signal-to-noise ratio (SNR). An ad hoc procedure for mitigating these array perturbations is presented, and simulations show that some performance improvement is achieved.