Target Detection Performance in a Clutter Environment Based on the Generalized Likelihood Ratio Test

We propose a method to estimate unknown parameters(e.g., target amplitude and clutter parameters) in the generalized likelihood ratio test(GLRT) using maximum likelihood estimation and the Newton-Raphson method. When detecting targets in a clutter environment, it is important to establish a modular model of clutter similar to the actual environment. These correlated clutter models can be generated using spherically invariant random vectors. We obtain the GLRT of the generated clutter model and check its detection probability using estimated parameters.