Array Gain And Phase Error Calibration In Direction Estimation

The high-resolution estimation algorithm haS been widely applied in radar,sonar,mobile communications and other fields in recent years.Both theory and practical application show that those performances are superior in the ideal array model.The peformances of these algorithms are perfect in ideal array model both in theory and practical application,but slowing done dramatically in error condition, even failed. As the restriction of manufacture-technics and construction-technique,array errors are often widespread.Most of the array errors can often be attributed to gain and phase errors in the ground array.So it's very important to study on DOA estimation algorithms under the conditions of the gain and phase errors in array.For the ubiquitous gain and phase errors in array,it's studed calibration methods of DOA estimation based on sub-space algorithm and improved artificial fish-swarm algorithm,Matching Pursuit algorithm in this paper.The main works of the thesis are as follows:1.The model of array and model of sensor gain and phase errors were constructed,an analysis for effect of sensor gain and phase uncertainty on spatial category of high-resolution algorithm is presented.2.Self-calibration methods of gain and phase errors for uniform linear array(ULA) sensor are designed;this method is base on the structural features of ULA covariance matrix and via iteration to calibrate errors,a better effect displayed.The method does not need the initial value of parameter,and it's easy to implement.Source-calibration methods is designed for direction finding with uniform circle array(UCA)when the sening of assistant source signal exist errors,this method reduce the need for precise assistant source signals,and effect of calibratiOn is well.3.Problem of array gain and phase errors and DOA joint estimation are resolved by combination of improved AFSA and MP algorithm,at the same time,self-calibration of gain and phase errors is tealized in ULA and UCA,and the methods are extended to DOA calibrating estimation of wide-band LFM signal.Those self-calibration methods do not need the initial Value of parameters,and have the greater range of errors calibration,be able to achieve accurate estimation of error parameters and angle in most cases.In addition the approachs have good performance in the case of serious array error and low SNR condition.