A Subspace-Aware Kelly's Detector using Reduced Secondary Data with Fast and Slow Time Preprocessing

Abstract

In homogeneous clutter environments, as the number of secondary cells used increases and eventually goes to infinity, the performance of the adaptive detectors increase and eventually goes to the performance of the max-SINR filter. However, this is not the case when the clutter is heterogeneous. In such environments, the adaptive detectors need to use small numbers of secondary cells. In such cases, reducing the dimension of the space in which the estimation occurs increases the detection performance. It is shown that reducing the dimension with the generalized eigenspace of signal and clutter subspaces instead of the conventional DFT subspace results in a remarkable increase in the detection performance and significantly reduces the number of secondary cells needed in the detection process. Combining this operation with a proper fast-time preprocessing method, a fast and robust detector is proposed. Simulation results are provided for comparing the detector's performance with other conventional radar detectors.