Abstract
In the future world, we require higher QoS in communication systems. How to imporve QoS efficiently with limited bandwidth has become an important issue. Recently, the concept of cognitive radio has emerged and is regard this as a solution to the limitation of bandwidth. There are cognitive users(CRs) and primary users(PUs) in a cognitive radio system.In this thesis, two kinds of users are in different communication system i.e. Digital TV and mobile system. Generally, two different communication systems need two different frequency bands for these two different systems so that we need lots of bandwidth to support them. However, in cognitive radio system these two systems can use the same frquency band as long as CRs will not cause interference to PUs. To null the interference, we used to adopt Transmit Beamforming technology with full CSI, but we cannot obtain full CSI in Cognitive Radio system since PUs and CRs are two different systems. In this thesis, we construct a Cognitive Radio system with Transmit Beamforming to reduce the interference between CRs to PUs with only DOA information without full Channel State Information(CSI). At first, MUSIC estimation algorithm is applied to estimate the Direction of Arrivals(DOAs), and then we utilize this space information to apply Transmit Beamforming. Moreover, KR-MUSIC algorithm is applied to alleviate the constraint that the number of antenna is less than the number of DOAs. The accuracy of our DOA estimation algorithm is shown by simulation, and then we verify Transmit Beamforming can indeed reduce the interference to PUs so that the BER would be lower. After verifying with simulation, in this thesis we have constructed a real-time Cognivive Radio system with Universal Software Radio Peripheral(USRP) and demonstrate its performance.