Cognitive radio has been recently proposed as a promising technology to improve the spectrum utilization. In this paper, we consider the coexistence between a large number of cognitive radio users and a licensed user in order to enhance the spectrum efficiency. With the deployment of M antennas at the cognitive base station, an efficient transmit beamforming technique combined with user selection is proposed to maximize the downlink throughput and satisfy the signal-to-interferenceand-noise ratio (SINR) constraint as well as limit interference to the primary user. In the proposed user selection algorithm, cognitive users who are nearly orthogonal to the primary user are first pre-selected so as to minimize the mutual interference. Then, M best cognitive users who are nearly mutual orthogonal to each other are scheduled from those pre-selected cognitive users. Simulation results show that our proposed method is able to achieve high sum-rate throughput, with affordable complexity. Moreover, our proposed technique with equal power allocation suffers a negligible performance loss compared to the one with the optimal power allocation.