A compressed sensing based design for formation of range-Doppler maps

Abstract

Compressed sensing (CS) based reconstruction methods have been in much focus over the last years as they provide a mean to work with limited amount of data. In this article we present an application of CS and sparse reconstruction in a radar setup to form range-Doppler maps. We characterize a radar system where instead of transmitting a train of pulses the pulse transmission occurs in a sparse manner. An array based radar may thus for example alternate beams between two or more angles within the same processing interval applying the entire array. We show that gaps in data can very adequately be filled in by sparse reconstruction which can also be used to extrapolate additional values. Comprehensive simulations covering various scenarios, including sea clutter conditions, are used to demonstrate the effectiveness of the scheme.