User-parameter-free robust adaptive beamforming algorithm for vector-sensor arrays within the hypercomplex framework

The major flaw of the conventional diagonal loading (DL) method is that it is unclear to choose appropriate DL levels or user-parameters (UPs), though several remarkable contributions have been made to regularize model errors without UPs. An UP-free algorithm for two-component vector-sensor arrays, which is robust to steering vector errors, is considered. The algorithm is within the hypercomplex framework using quaternions, and the optimal solution is found at the maximal correlation between the quaternionic and complex outputs. The performance of the proposed beamformer is illustrated via numerical simulations and is compared with several other UP-free adaptive beamformers.