BP-based matrix completion for near-field MIMO array imaging

The near-field array-based imaging radar systems have been widely used for concealed weapon detection, medical imaging, etc. However, traditional systems always require a large number of real antenna elements, which increase the cost and complexity of the systems. This paper firstly introduces an optimization method for near-field MIMO array design referring to the convolution principle. The antenna elements of the MIMO array are sparse which can obviously decrease the cost and complexity of the system. The back-projection (BP) algorithm is usually used for the MIMO radar imaging, which is a time-consuming operation. Thus, the paper presents a fast imaging algorithm which combines matrix completion with BP. Matrix completion is a new technique which can be applied to recover a low-rank matrix from a subset of the matrix entries. Simulations are provided to demonstrate the performance of the proposed algorithm, which proves that it is an effective way to solve the near-field sparse MIMO array imaging problem in a fast speed.