Adaptive beamformers are sensitive to model mismatch, especially when the desired signal is present in training snapshots or when the training is done using data samples. In contrast to previous works, this correspondence attempts to reconstruct the interference-plus-noise covariance matrix instead of searching for the optimal diagonal loading factor for the sample covariance matrix. The estimator is based on the Capon spectral estimator integrated over a region separated from the desired signal direction. This is shown to be more robust than using the sample covariance matrix. Subsequently, the mismatch in the steering vector of the desired signal is estimated by maximizing the beamformer output power under a constraint that prevents the corrected steering vector from getting close to the interference steering vectors. The proposed adaptive beamforming algorithm does not impose a norm constraint. Therefore, it can be used even in applications where gain perturbations affect the steering vector. Simulation results demonstrate that the performance of the proposed adaptive beamformer is almost always close to the optimal value across a wide range of signal to noise and signal to interference ratios.
Bibliographical noteFunding Information:
Manuscript received August 07, 2011; revised January 24, 2012; accepted March 14, 2012. Date of publication April 09, 2012; date of current version June 12, 2012. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Konstantinos Slavakis. This research was supported by the Israeli Science Foundation under Grant 1240/09.
- Covariance matrix reconstruction
- robust adaptive beamforming
- steering vector estimation