Research On Beamforming For Antenna Arrays

Wireless communication plays a more and more important role in people’s life. The negative outcome is that information security should not be ignored at the same time. In recent years, the combination of the physical layer security technology and the upper layer counterpart has attracted growing attentions, which enhances the security of wireless communications significantly. Beamforming, as an important skill in the physical layer security, is under the way of incessant improvement. Transmitted signals are weighted in an array, which leads to the signals having different gains in different directions. The beam pattern is expected to have high gains in the desired direction and nulls in the undesired directions. Therefore, the legitimate users can receive high SINRs (Signal to Interference plus Noise Ratio) while the eavesdroppers can only obtain low SINRs, which reaches the goal of QoS (Quality of Service) discrimination in different directions.The main works of the thesis is as follows.Firstly, when the quality of the wiretap channel is poor, only useful information is transmitted. Considering that the eavesdroppers can be either near the mainlobe or in the sidelobe of a beam pattern, the traditional nulling or polynomial representation algorithms are used to suppress signal energy in eavesdropping directions.Secondly, when the quality of the wiretap channel is good, even superior to the quality of the main channel, both useful and AN (Artificial Noise) are transmitted.With the useful information uniformly weighted, the weighted coefficients for AN can be obtained effectively by IDFT (Inverse Discrete Fourier Transform) weighted vector, least square error pattern synthesis, asymmetric pattern synthesis and the virtual interference pattern synthesis.Thirdly, in order to further improve power efficiency, AN together with transmit beam forming is jointly optimized. Assuming that the transmitter knows the perfect CSI (Channel State Information) of legitimate users and considering three cases that the transmitter can also know the perfect CSI,imperfect CSI or even nothing of the eavesdroppers, respectively, we propose to utilise CVX (Convex Optimization) to minimize the total transmit power subject to a target SINR constraint on legitimate users and limited SINR on all eavesdroppers.Finally, the transmitter transmits specific interferences instead of formerly designed random AN. Considering that the solution is non-convex, GA (Genetic Algorithm) is used to deal with this problem. Simulation results show that the proposed beamforming scheme is more power efficient than transmitting random disturbance, in discriminating the eavesdro- ppers’ reception performance.