Performance Comparison Between Statistical-Based and Direct Data Domain STAPs

In this paper the performances of two space-time adaptive processing (STAP), namely statistical-based methods and the direct data domain least squares (D3LS) method, are compared. The number of secondary or training data needed in the processing is a major concern in many applications. In the situation that the radar platform moves relatively fast, the smallest number of secondary data is preferred. While the statistical-based methods require training data in estimating the interference covariance matrix, the D3LS does not. The D3LS utilizes the interference information in a space-time snapshot under test itself in canceling this interference. The goal of this paper is to evaluate the performance of these two approaches by comparing the accuracy of the signal amplitude estimation when the number of training data is varied