Abstract : Array signal processing is a key area in signal processing, used in numerous classical and modern applications such as radar, sonar, wireless communication, acoustics, robotics, smart cities, and autonomous vehicles. A critical component of many array processing methods is the spatial correlation matrix of the array-received signals, which holds important spatial information. Typically, these matrices are used based on their positive-definite structure for beamforming, spectral analysis, and optimization. However, most methods treat these matrices as if they are in a simple Euclidean space, not taking advantage of the proven fact that they actually lie on a more complex Riemannian manifold.