Practical joint domain localised adaptive processing in homogeneous and nonhomogeneous environments. 2. Nonhomogeneous environments

For pt.1 see ibid., vol.147, no.2, p.57-65 (2000). The second part of this two-part paper deals with space-time adaptive processing (STAP) in nonhomogeneous environments. This paper introduces a new two-dimensional nonstatistical, direct data domain (D3) STAP algorithm and a hybridisation of this D3 approach with the joint domain localised (JDL) algorithm. The D3 algorithm replaces the non-adaptive transform used in Adve et al. (1999) (part 1) thereby suppressing discrete interference. A second stage of statistical processing in the angle-Doppler domain suppresses residual correlated interference. This new two-stage hybrid STAP technique allows for the application of the JDL algorithm within range cells determined to be nonhomogeneous. The development here draws heavily on the formulation presented in Part 1 of this paper. The work presented brings together two different aspects of STAP research: statistical and D3 processing. In doing so, this research fulfils an important need in the context of practical STAP, particularly knowledge-based algorithms. The envisioned system uses multi-pass processing to determine key information regarding the interference scenario. Depending on the homogeneity of the interference, the algorithm in Part 1 or that in Part 2 of the paper may be used