Traditional methods for beamforming can be grouped in two families, the time reference beamformer and the spatial reference beamforming. Nevertheless, the increasing demand on receivers for location systems and high bandwidth over frequency selective fading able to properly manage multipath and co-channel interference motivates the need for versatile processing able to cope with both problems. This paper presents a new beamforming procedure, derived from a ML-like framework, which is able to either remove full coherent arrivals, as well as, to enhance them when required yet preserving the receiver architecture and low design complexity. The performance of the beamformer is also tested in the multiuser broadcast scenario.

Peer Reviewed