Modified Root-MUSIC Algorithm for Target Localization Using Nyström Approximation

Target recognition and localization in communication systems present significant challenges. Often, computationally expensive algorithms that use the entire sensor array are necessary. In this article, we propose a novel methodology that leverages spectral factorization to address these challenges. Using the Nyström method on a randomly selected subset of array sensor outputs, we construct an advanced fast-modified root-MUSIC algorithm. This approach enables the efficient detection and localization of multiple narrowband far-field sources with enhanced resolution and reduced computational complexity. Through extensive computer simulations, we demonstrate that our proposed method achieves estimation performance comparable to traditional root-MUSIC techniques while using only one-fourth of the sensor array outputs. By integrating spectral factorization and the Nyström method, our methodology effectively mitigates computational overhead, providing researchers with a compelling solution for precise target detection and localization in communication systems.