A simple, low-computation peak detection algorithm for the angle-of-arrival spectrum for signal subspace methods

The array manifold, which is a collection of array steering vectors associated with k calibrated directions of arrival, is a commonality for the class of signal subspace methods. In actual direction finding (DF) systems, the array manifold (calibration table) is usually generated by measurements in a noisy environment. Subsequently, the resulting angle-of-arrival (AOA) spectrum, corresponding to as-yet-undetermined signal directions, may be coarse rather than smooth. This results in a costly computational search for the desired peaks. A simple, low-computation solution to determining the true signal peaks for the AOA spectrum is presented