338 Chapter | 8 High-Resolution DOA Estimation with Higher-Order Statistics PD-6-MUSIC, UP-PD-6-MUSIC-1, and UP-PD-6-MUSIC-2, respectively, with performances computed from 300 realizations. Note the capability of PD-2q-MUSIC to process underdetermined mixtures of sources provided that P P max given by Equation (8.75) or (8.79). Note also the poor performance of 2q-MUSIC for the considered scenario due to the low input power of the weakest source at the sensor outputs. Better performance would be obtained for higher values of L. 8.8 CONCLUSION In this chapter, the philosophy, properties, implementations, both parallel and sequential, and performance of HO HR DOA estimation methods were presented through description of the 2qMUSIC and PD-2q-MUSIC algorithms families (q 1). These are well suited for any kind of array with identical and diversely polarized sensors, respectively, for several arrangements of the 2qth-order data statistics. It was shown that HO cumulant-based HR DOA estimation methods have many advantages over SO methods. They are asymptotically robust with respect to a Gaussian background noise with spatial coherence that is unknown, and, because of a virtual increase in both the effective aperture and the number of array sensors, they allow for increasing resolution, increasing robustness to mod-