Phase-stability detection of stationary targets

The SAR (synthetic aperture radar) detection of stationary targets hidden in highly cluttered environments by using phase stability is studied. The theoretical performance of such a detection scheme is evaluated using simplified target and clutter models. Detection is based on a model where targets contain a concentration of phase-stable scatterers, and clutter is made up of phase-unstable scatterers. The proposed algorithm involves the transmission of three equally spaced frequencies. Each pixel is characterized by its observed stability, a linear combination of the phase shifts at each of the three frequencies, whose weights are +1, -2, +1, respectively. A mask is a collection of pixels chosen to match the possible geometry of a target. The observed stability of a mask is the weighted average of the squares of the observed stability of the pixels it contains. Assuming that target masks contain a sizable number of phase-stable pixels and that clutter masks do not, the observed stability of a mask can be used to obtain detection statistics. Power should be sufficient for SNR of at least 0 dB/pixel for each frequency. Phase-stability detection is only marginally improved by substantially improved SNR. Multiple scatterers in a pixel destroy phase stability unles the dominant scatterer has at least twice the combined RCS of all other scatterers combined. Phase-stability in target pixels could also be degraded by other means, such as foliage cover, scattering nets, and radar absorbing materials. The method has one substantial advantage over an amplitude-based system: detection statistics are independent of clutter strength.