Abstract
An interference suppression algorithm is proposed for canceling the spatially correlated background noise and interference in EEG/MEG source localization applications. Rather than using the standard prewhitening approach, the proposed algorithm attempts to directly null interference using a projection operator obtained from a set of secondary, control-state data. The use of direct interference cancellation is shown to be significantly more robust than prewhitening, due to the elimination of the need for computing covariance matrices, and the relaxation of the assumption of temporal stationarity. Simulation examples are presented to demonstrate the robust performance of the algorithm. ©2010 IEEE.