Study on the performance of DOA estimation algorithms

In this paper, the signal model of ULA is presented, and then two kinds of Direction Of Arrival (DOA) algorithms, traditional MUSIC [1-2] and spatial smoothing MUSIC [3-5], are analyzed with comparison with DOA using co-prime arrays based on convex optimization [7]. Finally, spectra and RSME performances are compared. Traditional MUSIC algorithm requires received signals to be incoherent, while spatial smoothing MUSIC algorithm can decorrelate the incoming coherent interface signals, and DOA using co-prime arrays can achieve the detection of more targets with fewer antennas than the other two algorithms. The simulation result shows that traditional MUSIC has higher accuracy when processing incoherent signals and its RSME is larger than spatial smoothing MUSIC processing coherent signals. Compared with the spatial smoothing MUSIC algorithm, the Root Mean Square Error (RSME) of DOA using co-prime arrays is smaller, and its DOF is larger, while the computational complexity is higher.