Comparative Analysis of Fourth Order Cumulant Based ESPRIT Algorithms

Higher order statistics based subspace methods are extensively used for Direction of Arrival (DOA) estimation. This paper compares different versions of fourth order cumulant based ESPRIT algorithms for DOA estimation in terms of accuracy, resolution and computational complexity. The popularity of cumulant based methods is because of their better resolution and their ability to perform well even in the case of high noise correlation. To further improve the accuracy and resolution, recently multiple invariance based cumulant ESPRIT methods are developed. However, they are far more computationally expensive which poses a question on their utility. A simple directive approach is proposed in order to reduce the computational complexity for the case of multiple in variances. It is observed that directive approach gives better results in terms of computational time and accuracy. Computer simulations are presented to compare all the algorithms in consideration.