Research On Anti-jamming Techniques Based On Adaptive Array Antennas

Array antennas play an important role in the current anti-jamming techniques. Based on certain criteria and algorithms, they can adaptive adjust the weights of the elements, and as a result, the main lobe of the beam pattern can point to the desired signals and at the same time, the null or low side lobe to the interference signals. Therefore, the interference signals can be suppressed since the array antennas make them null. Beamforming algorithms are one of the main issues in array antenna research filed. Both covariance matrix estimation error which caused by infinite sample and the existence of coherent signals can lead to severe performance degradation. High computational complexity also limits their applications of the current algorithms; therefore, study of adaptive null beamforming algorithms with better anti-jamming performance is necessary.Some basic concepts of array antennas technique and its advantages and practices are firstly described in this dissertation, and the system structure and working principle are introduced and the received signals model is set. Secondly, some classical adaptive null beamforming algorithms, such as LMS, RLS, and DMI are introduced and some analysis and performance comparison are studied. Thirdly, this dissertation focus on the eigenstructure based beamforming (ESB) algorithm. Since conventional ESB algorithm using eigenvalue decomposition (EVD) brought about high computational complexity, this dissertation proposes an improved subspace approximation algorithm based on the research of existing subspace approximation algorithms. Furthermore, ESB algorithm can not form effective null steering beam to the interferers when coherent signals existed. This dissertation studies the Forward/Backward Spatial Smoothing technique based ESB algorithm (FBSS-ESB), and discusses a Lower-Rank Approximation based ESB algorithm (LRA-ESB) and an improved ESB algorithm using SVD based Toeplitz matrix reconstruction method (SVD-Toeplitz-ESB). Simulations with MATLAB verify the improved algorithms. Finally, a summary is made for the whole dissertation.