Recursive implementation of GLRT-based radar target detection

摘要

We consider recursive implementation of the natural frequency-based radar target detection for an augmented data vector. In the previous study, it was shown that, the probability of detection can be calculated using the probability density function (PDF) of the non-central chi-square distribution and that non-centrality of the chi-square distribution is dependent on the eigenvectors of a matrix. The essential idea in this paper is that the eigenvectors of the augmented matrix can be recursively calculated without computationally intensive eigendecomposition. To do that, we make use of how the QR factorization of the row-augmented matrix can be updated from the QR factorization of the original matrix to get the probability of detection recursively. The recursive formulation is validated by comparing the detection performance using the recursive method with that using non-recursive method.