摘要: Diagonal loading is one of the most widely used and effective methods to improve robustness adaptive beamformers. The authors consider its application case steering vector errors, i.e. when there exists a mismatch between actual interest presumed one. More precisely, problem addressed that optimally selecting level with view maximising signal-to-interference-plus-noise ratio in presence errors. First, an expression derived for optimal given error it shown this negative. Next, random errors are considered averaged respect probability density function yielding very simple average loading. Numerical simulations attest validity analysis show diagonal factor provides performance close optimum.