Amplitude And Phase Errors Calibration Of High-Frequency Over-the-Horizon Radar Array Based On AIS Information

The high frequency surface wave radar(HFSWR)is an effective means to observe the dynamic environment of the sea surface and the targets on the sea.Its monitoring range is generally up to 200 kilometers.It is necessary to know the array manifold accurately when using HFSWR to locate and estimate the current and ship targets.However,in practical engineering application,due to the difference of hardware itself and the change of complex background environment,the amplitude and phase characteristics of the receiving channels will appear a certain degree of deviation and disturbance,the array manifold will be distorted,and the performance of the direction finding algorithm will be significantly reduced or even invalidated.Therefore,the calibration of array amplitude and phase errors is one of the essential standard functions of HFSWR,and the research of related methods is also the key to the continuous improvement of radar performance.Based on the calibration method of single auxiliary source amplitude and phase errors,according to the basic principle of array signal processing,this paper proposes a calibration method of HFSWR array amplitude and phase errors,which combines a large number of ship target information provided by Automatic Identification System(AIS)and cluster analysis,and carries out comparative verification.The main contents of this paper are as follows:(1)Analysis of the array amplitude and phase errors and comparison of the commonly used array amplitude and phase errors calibration algorithms.The influence of array amplitude and phase error on the direction finding performance of radar system is analyzed by using the mathematical model of array received signal and numerical simulation.This paper introduces the working principle of three classical algorithms,which are active calibration based on single auxiliary source,self calibration based on subspace orthogonal and self calibration based on auxiliary array element.The performance of the three algorithms is compared by computer simulation.In addition,the direct wave calibration in active calibration and the sea echo calibration in passive calibration are studied.(2)Amplitude and phase error calibration of HFSWR array based on AIS information.The ship providing AIS information can be used as a calibrator to calibrate the amplitude and phase errors of radar array.However,it is often inaccurate to determine the ship echo spectrum points in the radar echo spectrum based on AIS information,which makes it impossible to calibrate the amplitude and phase of the radar array using a single ship AIS information.When AIS is rich in information,more ship echoes can be estimated in radar echo spectrum,and then multiple sets of radar array amplitude and phase error calibration results can be obtained.According to the fact that the correct calibration values obtained by different ships must be consistent with each other,a method is proposed to extract the correct calibration values from the error set obtained by using AIS information to estimate the amplitude and phase errors of radar array by using correlation coefficient analysis or kmeans clustering analysis,which makes the calibration method of radar array amplitude and phase errors based on AIS information applicable in engineering.(3)Calibration analysis of HFSWR measured data.In this paper,the proposed method is applied to HFSWR data processing.The 24-hour phase standard deviation of all antenna calibration values of the array is not more than 9%,and the 24-hour amplitude standard deviation is not more than 1d B.The DOA estimation error of direct wave is reduced from 20 ° before calibration to no more than 5 °.The results of ocean current inversion show that after calibration,the sea current coverage of radar observation is increased by about 15%,compared with that of buoy,the root mean squared error(RMSE)of the deviation between 24-hour radar observation and buoy observation is reduced from 18.3cm/s before calibration to less than 10.9cm/s after calibration,the correlation coefficient reaches 0.95.The results of the study effectively verify the accuracy and feasibility of the new method in the field data processing.