Robust direct data domain processing for MTI

Direct data domain space-time adaptive processing (D3-STAP) is an interesting approach for multi-channel radar moving target indication. The main difference with respect to conventional stochastic STAP resides in the capability to cancel the interference (both jammers and clutter) using only the information contained in the single range gate under test. Therefore, being implicitly a single snapshot interference cancellation technique, D3-STAP shows several advantages compared to stochastic STAP in fast varying interference scenarios, where the availability of secondary data for interference statistics estimation is limited. In this chapter, a novel approach for D3-STAP is described. This amelioration overcomes the main limitation of D3-STAP in its classical derivation. In fact, target detection performance of D3-STAP is severely deteriorated in case of inaccurate knowledge of target parameters, namely direction of arrival (DOA) and Doppler frequency. To overcome this problem, a robust D3-STAP (viz. RD3-STAP) implementation is shown which takes into account a possible mismatch between nominal and actual target parameters. The approach reformulates the D3-STAP problem in the context of convex optimization, and it can be applied to the different implementations of D3-STAP, namely forward, backward and forward-backward methods. In addition to that, an implementation of RD3-STAP with dimension reducing transformations is shown which limits the number of degrees of freedom. The effectiveness of the proposed approach is shown both in simulated scenarios and by direct application to real data taken from the experimental multi-channel radar system Phased-Array Multi-functional Imaging Radar (PAMIR) developed at Fraunhofer-Institute for high frequency physics and radar techniques; in German: Fraunhofer-Institut fur Hochfrequenzphysik und Radartechnik (FHR). Finally, possible applications of RD3-STAP to multi-channel synthetic aperture radar and to target DOA estimation cases are presented.