A multichannel auto-regressive GLR detector for airborne phased array radar applications

This paper considers the Gaussian multichannel binary detection problem. A multichannel generalized likelihood ratio (GLR) is presented using a model-based approach where the signal and its amplitude are known and additive noise is characterized by autoregressive (AR) vector processes which can be a suitable model for ground clutter in most scenarios involving airborne surveillance phased array radar. The detector can use secondary vectors to improve its performance. Although, in our model, the amplitude of the signal must be known, but according to its excellent performance, an approximation of the GLR on the amplitude is also presented. The detector performance in various conditions and the result of comparison with Kelly's GLR and adaptive matched filter (AMF) detectors is presented by computer simulations.