The impact of inverse matrix computation on the results of the Capon algorithm

Algorithms for estimating the direction of signal arrival are crucial in various applications, particularly in acoustics. They are essential for accurately locating sources of both desired and undesired sound signals. This research focuses on the performance analysis of the Capon algorithm using an approximated inverse covariance matrix. Instead of using direct methods for matrix inversion, iterative methods are employed, potentially leading to shorter processing times and reduced memory usage. Furthermore, the performance of these iterative methods is compared to the direct Cholesky method in terms of computational operations.