Matrix completion-based MIMO radar imaging with sparse planar array

Abstract

Multiple-input multiple-output (MIMO) radar with sparse planar arrays is expected to provide one snapshot imaging of complex motion targets at low hardware costs. In recent years, compressive sensing (CS) has been applied to MIMO radar imaging with limited antenna arrays. However, CS has to convert the matrix into a vector, which results in a large measurement matrix and a demanding amount of storage. Matrix completion (MC) is a new theory of recovering the entire data with partial observations. Compared with CS, MC is more suitable for matrix operations and avoids the basis mismatch problem. Due to the potential advantages, we introduce MC to MIMO radar imaging using a sparse planar array. Despite the success of MC in other fields, direct application in MIMO radar imaging leads to several problems, such as strict restrictions of the low-rank property and uniform random samples in MC. To solve these problems, a structured MC(StMC)-based method is proposed. With the method, MIMO radar imaging via a sparse planar array is converted to complete a low-rank structured matrix, which can be easily solved as a semi-definite program problem. Experimental results confirm the validity of the proposed method.