Enhanced Uniform Linear Array Performance Using a Modified MVDR Beamforming Algorithm in a Multi-Signal Environment

For a uniform linear array (ULA) in a multi-signal environment (MSE) with one desired signal, several interfering signals and noise, which are assumed to be mutually uncorrelated, an adaptive null-forming beamformer is presented in this paper. The proposed method is based on the minimum variance distortionless response (MVDR) beamformer, and adds a zero forcing mechanism. In particular, the steering vector of the desired signal is projected onto the subspace orthogonal to the space spanned by the interferer steering vectors, and the vector resultant is used as the steering vector in the MVDR solution. The main idea is to make the nulls deeper in the directions of the interference and beam the mainlobe more accurately in the direction of the wanted signal. In the ideal covariance model, we prove analytically that the proposed weight vector is exactly the weights that satisfy both the distortionless and null constraints. The method was tested using simulations with an eight-element ULA half wavelength apart with a desired signal at 0°, interferers at 14.5° and 40.5° and input SNR and INR of 10 dB. The number of snapshots ranged from 1,000 down to 7. Averaged over three runs, the proposed beamformer achieved an output SIR approximately 33 dB higher than that of the conventional MVDR beamformer at 1,000 snapshots and approximately 54 dB higher at 15 snapshots. With fewer snapshots, the mainlobe of the conventional beamformer shifted away from the desired direction, while that of the proposed beamformer remained closely aligned with it.