Research On Beamforming Algorithms And Applications Of Array Antennas

Array antenna beamforming has a wide range of applications,such as radar,sonar,wireless communications,etc.Its main functions include: filtering the airspace,estimating the direction of the signal arrival,creating conditions for the location of the signal source and so on.As a typical algorithm,Capon beamformer has attracted abundant attentions due to its good performance and flexible expression.However,this algorithm can achieve good performance only under certain conditions such as assumptions of array calibration.There are some mismatches in practical application which seriously affect the performance of the algorithm.Therefore,to improve its robustness,lots of improved algorithms have been proposed.In this dissertation,three new algorithms are proposed to improve the performance of Capon beamformer.To get the optimal loading factor,a constrained optimization problem is designed in most existing diagonal loading algorithms,and then multiple iterations are performed with the help of optimization software.This calculation process is often complicated.Two improved diagonal loading algorithms based on the sample covariance matrix are presented in this dissertation.The proposed algorithms do not need any optimization software and have good effect on robustness and adaptability.Furthermore,compared with other reference algorithms,the proposed algorithms have lower computational complexity.Simulation results show that the proposed algorithms are with better robustness in the presence of signal steering vector mismatch caused by deviations such as the direction of observation.The existing algorithms of reconstructed sample covariance mostly adopt the idea of spatial power spectrum integration.By modifying the integral interval,they can improve the accuracy of reconstructed covariance matrix at the expense of high computational complexity,a new algorithm based on covariance matrix reconstruction is proposed in this dissertation.In the proposed algorithm,the sample covariance matrix is processed to remove the desired signal components,and then the covariance matrix is reconstructed.In addition,the steering vector of the desired signal is estimated with the help of the parallelogram rule of the vector.Through the above processing,the performance of the proposed algorithm is significantly improved.Finally,computer simulation results further demonstrate the effectiveness and superiority of the proposed algorithm.