In synthetic aperture radar theory, periodic spatial sampling that satisfies Nyquist theorem can be used to generate imagery with minimal ambiguities. A two-dimensional (2-D) grid of uniformly spaced aperture samples can be used to generate three-dimensional (3-D) radar imagery. However, 2-D apertures typically result in an untenable number of samples for practical implementation. The spacing between aperture samples can be increased to reduce the number of samples at the potential cost of introducing ambiguities. Since the sampling is uniform, this can introduce grating lobes within the image area. Grating lobes are erroneous points of coherence that result from sub sampling (i.e., not satisfying Nyquist theorem) a periodic array. The recursive sidelobe minimization (RSM) algorithm removes sidelobes by exploiting the varying null positions in images formed with random subapertures. However, grating lobe spacing is generally unaffected by subaperture selection in periodic arrays. This paper presents a modification to the RSM algorithm which removes grating lobes by randomizing the operating center frequency for each iteration of the algorithm.
|