Root-MUSIC Algorithm MATLAB Implementation

In modern pulse radar systems, additive noise frequently presents as colored noise, which causes the traditional MUSIC algorithm to fail despite its super-resolution capability. To address this limitation, we propose an enhanced method - the Root-MUSIC algorithm. This algorithm employs polynomial root-finding techniques to replace the spectral search process in conventional MUSIC. The implementation involves constructing the signal covariance matrix from received data, performing eigenvalue decomposition to separate signal and noise subspaces, and formulating a polynomial whose roots correspond to direction-of-arrival estimates. Root-MUSIC demonstrates superior performance in colored noise environments by effectively handling the noise covariance structure, thereby improving radar detection accuracy and precision. The MATLAB implementation typically includes functions for array signal processing, root calculation using the 'roots' function, and angle estimation through phase extraction from complex roots.