Experiment design-based regularization technique for scattering inverse problem solution in radar imaging

The crucial theoretical aspects in solving scattering inverse problems, as required for radar sensing and image formation, are related to the development of system-oriented statistical signal processing techniques, exploiting in some optimal way the available model-level and system-level "degrees of freedom". Control of these degrees of freedom is the hope for solving many existing algorithm design and system-level problems that cause the ill-posed nature of the radar image formation inverse problems. We address a new approach to radar imaging (RI) problems, stated and treated as ill-posed problems of restoring the extended object power scattering function by processing the data signals distorted in the stochastic measurement channel. By exploiting the idea of combining the experiment design and descriptive regularization theory methods (see Falkovich, S.E. et al., Radio i Sviaz, 1989; Harmanci, K. et al., IEEE Trans. Sig. Proc., vol.48, no.1, p.1-13, 2000; Kravchenko, V.F. et al., J. Commun. Technology and Electronics, vol.45, no.8, p.872-5, 2000) for the scattering inverse problems solution, we propose a fused experiment-design-regularization (EDR) technique for high-resolution radar image formation. With this technique, we derive a family of EDR imaging algorithms and illustrate their efficiency with simulation results.