Direction of Arrival Estimation on Sparse Arrays Using Compressive Sensing and MUSIC

Now-a-days radar target detection is a safety critical system in automotive applications involving autonomous driving. The security and trustworthiness of such systems can be improved using a sparse interleaved array. In this paper, we present a new method for target detection for this sparse array, which combines the root-MUSIC algorithm with orthogonal matching pursuit (OMP) to estimate the azimuth angle and recovers the signal. It compensates the major drawbacks of compressed sensing, in which the choice of underlying grid for azimuth angles affects the optimality of the solution. The elevation angles is then estimated from the approximated signal using root-MUSIC. Based on experimental results with 77 GHz MIMO system, we may achieve a high precision in angle estimation without aliasing as well as increased dynamic range of 3dB in comparison to conventional methods.