Research On Digital Beamforming Based On Signals Cyclostationarity For Smart Antenna

Many signals in communication and radar fields are cyclostationary. Recently, significant researches on blind adaptive beamforming have been carried out by exploiting the cyclostationarity of signals, the common advantages of these methods can be summarized as follows: only the cycle frequency of the desired signal is required, no reference signal is required, no advanced knowedge of the correlation properties of noise and interference is needed, and no complicated procedure for array calibration is necessary. Therefore these methods are the genuine blind adaptive beamforming methods.In this thesis, a series of researches on the blind adaptive beamforming based on the cyclostationarity of signals have been carried out, includingFor one thing, the philosophy on digital beamforming and the feature of the cyclostationary signals are introduced, the advantage and disadvantage of such algorithms as CAB/CCAB/ECAB are evaluated in detail.And for another, to resolve the problem of ECAB algorithm that intensive computation and unrealtime processing in updating the signal subspace, an improved algorithm for ECAB is presented. The improved algorithm select 'Unconstrained Optimimum Gradient Algorithm' to track subspace after comparing the performance of three subspace tracking algorithms. Because computation complexity of the selected subspace tracking algorithm is lower, the algorithm update subspace more accurate than the proposed algorithm before, while convergence is also good, the analysis shows that the improved algorithm can increase output SINR and suppress interference more effectively, its computation is small and its convergence is rapid. The simulation experiments confirm the improved ECAB algorithm. The performance of the improved ECAB algorithm will help the algorithm realtime implemention for application system.In addition, to resolve the problem of CCAB algorithm that the correlation matrix is suitable for stationary signals or approximative stationary signals and not suitable for cyclostationary signals, an improved algorithm for CCAB is proposed. The improved algorithm use exponential temporal window instead of...