Optimization of a Sparse Aperture Configuration for Millimeter-Wave Computational Imaging

We present two techniques for optimizing the position of transmitter and receiver modules on a sparse aperture for a millimeter-wave computational imaging system. The first technique uses an easily computable spatial representation of the transmitter and receiver array, called the coarray, to ideally distribute the spatial frequency components probed by the imaging setup. The second approach involves maximizing the information added by a complete measurement of the scene by the system. This approach is analogous to the system capacity maximization frequently employed in wireless communication. We show that employing aperture configurations optimized using these two techniques over commonly used standard aperture configurations results in 30% less mean-squared error when used to reconstruct a particular ensemble of 400 arbitrary 2-D objects. Finally, we discuss the similarities and differences between the two optimization strategies in terms of imaging performance and computational speed, including a case when one strategy performs better than the other.