Space-time adaptive processing for forward looking arrays

The paper applies space-time adaptive processing (STAP) to forward looking arrays. Traditionally, STAP research provides results for a side-looking array where a linear relationship exists between clutter Doppler and azimuth sine. This relationship does not hold in the forward looking case. Additionally, mainbeam clutter in the side-looking boresight case is range independent and, consequently, the clutter notch is always at zero hertz. In contrast, mainbeam clutter is range dependent in the forward looking case, resulting in multiple clutter notches for range ambiguous clutter. This range dependency corrupts the independent, identically distributed assumption required for the samples used in covariance estimation. The assumption can still be applied for a small data set, e.g., small range extent. Therefore, partially adaptive techniques with fewer degrees or freedom and correspondingly smaller sample support requirements must be used. Results are presented in the form of clutter power spectral density, signal-to-interference-plus-noise ratio, improvement factor, antenna patterns, and detection probability.