Higher-order time frequency-based blind source separation technique

This letter considers the separation and estimation of independent sources from their instantaneous linear mixed observed data. Here, unknown source signals are estimated from their unknown linear mixtures using the strong assumption that the sources are mutually independent. In practice, separation can be achieved by using suitable second- or higher-order statistics. The authors propose a novel source separation technique exploiting fourth-order time frequency distributions. A computationally feasible implementation is presented based on joint diagonalization of the matrices of the principal slices of time-multifrequency domain of support of the cumulant-based Wigner trispectrums. A numerical example demonstrates the effectiveness of the proposed approach.