Adaptive and Distributed Beamforming for Cognitive Radio

Abstract

Cognitive Radio (CR) is an energy efficient technique that is capable of optimizing the premium radio resources, such as power and spectrum. In this thesis, we focus on exploiting spatial diversity for CR. We have adopted two spatial signal processing techniques, i.e., Adaptive Beamforming (ABF) and Distributed Beamforming (DB) for CR users and CR networks, respectively. We have investigated and proposed a Bayesian ABF technique to CR Base Station (BS), which is able to direct CR BS main beams to CR users even when the Direction of Arrival (DOA) of each CR user is uncertain or completely unknown. Consequently by using this method, a CR BS at uplink (receiving) can adaptively enhance the signals of CR users by directing its main beams towards their wanted directions. We have also developed a Null Broadening (NB) technique for CR BS to be capable of generating spread nulls in the beampattern around directions of Primary Users (PUs), which guarantees radiated power reduction towards directions of PUs even considering scattering and multipath effects. We have introduced a DB method to distributed CR networks, which are constituted of distributed CR nodes. Two multi beam generating methods have been presented for DB to generate more than one main beam towards Distant CR (DCR) users. Due to the limitation of the DB method, the main beam of the CR network will be extremely narrow considering the possible CR working frequency. Therefore we have proposed a Nodes Selection (NS) method to select proper CR nodes from the CR network to perform DB, and thus the beampattern has a wider main beam while maintaining the lower sidelobe levels towards PUs.