Research On Robust Adaptive Beamforming Algorithm

Adaptive beamformer is an important part of the array signal processing field.Its main function is to align the main lobe with the direction of the incoming wave of the desired signal,and then form a null in the position of the interference signal to achieve the role of spatial filtering,but in practice There are a variety of error factors in the application that cause large errors in the steering vector and the sample covariance matrix,resulting in a sharp drop in the performance of the beamformer.To counter the effects of these errors,it is of great significance to study robust adaptive beamforming algorithms.The main work and innovations of this paper are as follows:(1)In view of the error caused by various error factors to the steering vector,this paper proposes an algorithm of correcting the steering vector which based on a new constraint to improve the performance of the beamformer.This algorithm mainly constructs the orthogonal subspaces of the positive vector which belong to desired signal.The subspaces are then constructed and the constraint equations are constructed by the relationship between them to solve the error vectors.The method to choose the proper orthogonal subspace is given.(2)Aiming at the error of steering vector,this paper also proposes a method to solve the steering vector based on constructing constraint conditions of uncertain set.within the desired signal angle range based on the geometric relationship between the error vectors and the different angle steering vectors.The boundary value is constructed to obtain the constraint equation of the uncertain set,and then the convexity optimization of the objective function is solved to obtain the error vector.The algorithm can effectively counteract the distortion of steering vectors caused by various common error factors.(3)Aiming at how to effectively reconstruct the INC matrix,this paper proposes a method based on new signal source power estimation to reconstruct the INC matrix and a method to reconstruct the INC matrix based on the interference signal subspace.The first algorithm is mainly based on the geometric relationship between the signal subspaces of the eigenvalue decomposition which of the received covariance matrix and the signal steering vector.The expression of the signal power is obtained by derivation..The power estimated method has a low complexity and the performance is close to the optimal value.(4)This paper also proposes a method of reconstructing an INC matrix based on interference signal subspace.This algorithm uses the Capon power spectrum to obtain the steering vector of the interference signal,and then constructs a subspace orthogonal to the interference signal to reconstruct the INC matrix.The algorithm is essentially to find the vector with the largest coherence with the desired signal from the subspace orthogonal to the interference signal as the weighted vector.The biggest advantage of this algorithm is that it does not need to estimate the power of the signal which with less complexity,but only for higher INR scenarios.