Study On Robust Adaptive Beamforming Technology For Wideband Null Broadening

When the adaptive antenna array encounters sidelobe motion interference,array platform vibration,etc.,the null formed by the conventional beamformer is narrow,and the update speed of the adaptive weight vector cannot keep up with the change of the direction of the interference.This will cause the null position of the beamformer to be unable to align with the interference direction,and the interference suppression performance will be severely degraded or even completely invalidated.Therefore,designing a robust null-broadening adaptive beamformer that can form wider nulls in the interference direction and still have good interference suppression performance under non-ideal conditions is one of the goals of robust beamforming research today.In this paper,a robust null broadening adaptive beamformer is designed around narrowband signals and wideband signals.The main work is as follows:1.In view of the problem that the widening effect of the existing null broadening algorithm is easily affected by the diagonal loading(DL)value,two robust null broadening algorithms are proposed.The first proposed projection transform,constraint matrix reconstruction and diagonal loading null broadening beamforming(PCLNBB-1)algorithm builds a correlation matrix from the steering vector of the widened angle interval,and performs feature projection on the received data to obtain the reconstructed covariance matrix and constraint matrix of the received data.Aiming at the problem of large amount of calculation in the PCLNBB-1algorithm,the PCLNBB-2 algorithm is proposed.This algorithm uses the desired signaloriented vector projection instead of the constraint matrix reconstruction.Compared with the PCLNBB-1 algorithm,it avoids repeated inversions and reduces the amount of calculation.Simulation experiments show that under non-ideal conditions,the output signal to interference and noise ratio(SINR)of the proposed algorithm is better than other null broadening algorithms,and the effect of null broadening is not affected by the selection of diagonal loading values.2.Aiming at the problem that the existing wideband adaptive null broadening algorithm is not ideal for suppressing sidelobe motion interference,and the performance is general under non-ideal conditions,four robust wideband adaptive null broadening algorithms are proposed.The first proposed projection and diagonal loading wideband null broadening beamforming with frequency-invariant constraints(PDWNBB-FI-1)algorithm in this paper builds a correlation matrix from the stacked steering vectors of the null broadening interval to directly project the received data of the array to achieve the purpose of reducing the expected signal component and obtain wide null.At the same time,the algorithm reduces the number of constraints and maintains the consistency of the main lobe by adding response variation(RV)constraints.The second PDWNBB-FI-2 algorithm projects the received data of the subband array.The PDWNBB-FI-1 algorithm and PDWNBB-FI-2 algorithm have similar null broadening effect,but their performance is slightly different under non-ideal conditions.In order to avoid the desired signal in the received data of the array being regarded as interference suppression,the projection and covariance matrix reconstructing wideband null broadening beamforming with frequency-invariant constraints(PRWNBB-FI)algorithm is proposed.Unlike the previous two algorithms that project the received data,the PRWNBB-FI algorithm projects the stacked interference covariance matrix(SICM)removes the desired signal component,and have better null broadening results.Aiming at the complex structure of the covariance matrix and the large amount of calculation in the existing stacked interference-plusnoise covariance matrix(SINCM)reconstruction algorithms,the eigenvalue covariance matrix reconstructing wideband null broadening beamforming with frequency-invariant constraints(ERWNBB-FI)algorithm is proposed.The ERWNBB-FI algorithm uses the interference estimated value as the power of the virtual interference source instead of the power spectral density estimation of the subband.Compared with the existing broadband adaptive null broadening algorithm,the null broadening effect is the best,and the calculation amount is the lowest among the existing SINCM reconstruction algorithms.Finally,the simulation analyzes the related parameters of controlling null depth and main lobe width in the algorithm.The four algorithms mentioned in this article are compared with existing algorithms.The experiment verified that the four broadband adaptive zero-trap and width algorithms mentioned in this article have a better zero trap and width effect.They can form a wider null in the direction of the interference,which can effectively suppress the motion interference of the sidelobe.In nonideal conditions,the four algorithms mentioned in this article have a good effect of null broadening,while the output of the beamformer is better than the existing wideband adaptive null broadening algorithms.