A Novel Self-Calibration Technique for Linear Array Based on Modified MVDR Adaptive Beamformer

Adaptive smart antenna is a combination of antenna arrays and beamforming algorithm. One of the important adaptive beamforming algorithms is Minimum Variance Distortionless Response algorithm (MVDR), which has optimal weight and very good output Signal-to-Interference Noise Ratio (SINR). MVDR is very sensitive to the steering vector errors that may occur due to manufacturing or calibration errors at antennas of the array. Signal suppress (selfnull) phenomena appeared due to this error, where mismatch between presume and actual signals Direction of Arrival (DOA) occurred. This error also may be occurred during the work time of smart antenna and after the final calibration test due the hard weather changes. According to this weather change, the shape of array changes and leads to error in the antenna interdistance. This leads to major degradation in the MVDR beamformer performance. In this paper, a novel diagnoses algorithm of smart antenna self-calibration is proposed. This algorithm aimed to overcome the calibration/manufacturing or shape distortion errors that may essentially exist or may occur by weather changes effect. According to this algorithm the reference element is re-localized to be in the middle of the Uniform Linear Array and the rest antenna elements distributed identically at the two sides of this reference. This configuration generates two identical sub-arrays at the two sides of the reference. Due to identical performance of the two subarrays, the inter-distance change between array’s antenna elements easily distinguish. Simulation results showed that the proposed algorithm has an effective response against steering vector errors since the desired signal is served by the main radiation beam while the interference is blocked by null.