Further study on space-time adaptive processing with sum and difference beams

Most space-time adaptive processing (STAP) approaches for airborne surveillance radars have two common problems: high production and maintenance cost resulting from their need for a large number of calibrated coherent receiver channels, as well as poor performance in nonhomogeneous clutter environments, as pointed out in Wang (1995). To solve the above problems, a new STAP approach, called /spl Sigma//spl Delta/-STAP, is described in Brown (1996), which uses only the main sum (/spl Sigma/) and difference (/spl Delta/) beams of an airborne radar for adaptive suppression of clutter in the joint angle-Doppler domain. This approach dramatically reduces system implementation cost, as it requires only two digitized channels, and is applicable to both phased array and reflector-feed apertures. It has been shown via simulation that /spl Sigma//spl Delta/-STAP can offer near optimal performance and requires much smaller sample support than other approaches, which is critical for successful operation in severely nonhomogeneous clutter. As a continuation of our effort on this approach (Wang 1996), this paper shows some new results on the /spl Sigma//spl Delta/-STAP approach, including the desired characteristics of the /spl Sigma//spl Delta/-beam patterns, comparison between the /spl Sigma//spl Delta/-STAP and /spl Sigma//spl Delta/-DPCA, and the evaluation of its performance for three different beams.