Comparative study of IQML and MODE direction-of-arrival estimators

We present a comparative study of using the iterative quadratic maximum likelihood (IQML) and the method of direction estimation (MODE) for direction-of-arrival (DOA) estimation of narrowband incident signals with a uniform linear array. The conditions for local convergence and the theoretical performance of IQML are derived. Although the DOA estimates obtained via MODE are asymptotically statistically efficient, we show that those obtained via IQML are almost always inconsistent and have (much) larger mean squared errors. We also show that the amount of computations required by IQML is usually much larger than that required by MODE, especially for low signal-to-noise ratio (SNR) and a large number of snapshots.