BLIND SYSTEM-IDENTIFICATION USING MINIMUM NOISE SUBSPACE

Developing fast and robust methods for identifying multiple FIR channe ls driven by an unknown common source is important for wireless commun ications, speech reverberation cancellation, and other applications. I n this correspondence, we present a new method that exploits a minimum noise subspace (MNS). The MNS is computed from a set of channel outpu t pairs that form a ''tree.'' The ''tree'' exploits, with minimum redu ndancy, the diversity among all channels. The MNS method is much more efficient in computation than a standard subspace method. The noise ro bustness of the MNS method is illustrated by simulation.