High range resolution (HRR) radar signatures of objects provide a 'downrange' one-dimensional mapping of radar cross section (RCS). In the figure, an HRR radar is illuminating the aircraft from the left, and the HRR signature that results at each aircraft pose angle is depicted at the bottom of the screen. Certain features, such as the twin tails of the aircraft can be observed over many pose angles.

Radar cross section (RCS). In the figure, an HRR radar is illuminating the aircraft from the left, and the HRR signature that results at each aircraft pose angle is depicted at the bottom. Certain features, such as the twin tails of the aircraft, can be observed over many angles.

However, it can be seen that much of the signature appears to completely decorrelate over very small changes in target pose angle. This is due to 'Glint,' that is, the rapid change in the phase interference pattern among the dominant scatterers in each range bin. The signatures also change as a function of pose angle due to range-bin migration, the anisotropic nature of the scattering centers, and shadowing, albeit much more slowly.

HRR target identification can be accomplished via the estimation of physical characteristics of the observed target and/or by pattern matching. The pattern matching approach requires that a library of representative signatures be stored for comparison against received HRR profiles. The rapid decorrelation of these signatures with changes in pose angle would seem to make the pattern matching approach very difficult indeed.

Fortunately, it has been found that though these signatures are not stationary, they are often cyclo-stationary over reasonably large changes in pose angle. This cyclo-stationarity permits us to create representative templates of these signatures for each aircraft of interest over sufficiently broad pose angles that the storage of the library and its real time search become feasible.