A New Joint Diagonalization Algorithm with Application to Blind Source Separation
Feng Wang, Bin Wu · 2012
In this paper, a novel iterative algorithm for the joint diagonalization of a set of real symmetric matrices is presented. The approximate joint diagonalization of a set of matrices is an important approach in many blind source separations (BSS) applications. By applying the least-squares criterion, a classical nonlinear least-squares cost-function of the BSS problem is obtained. An explicit method for optimizing the cost function is derived and compared with other techniques in the BSS context. Both simplicity and efficiency were achieved by applying the algorithm to BSS problems.