Abstract

Adaptive beamforming is mainly used for interference rejection and for resolution when the array steering vector\nis already known. And adaptive beamforming is signal processing system which combined a signal in a manner to increase\nthe signal strength in a particular direction . Sometime there are some mismatches between the true value and assumed\nvalue,whichcausesdegradation in adaptive beamforming technique. In this paper, we characterize a novel neural network\nmethod to robust adaptive beamforming. In the proposed algorithm ,it depend on the class of diagonal loading styles which is\nthree-layer RBFNN(radial basis function neural network) and in evident modeling of improbability in the required signal\narray response. In this algorithm, the list of the best weight vector is noticed as a mapping problem, which can be modeled\nwith the help of RBFNN with input/output pairs.Here best performance under good conditions, provides the robustness\nagainst the signal steering vector whichreduce themismatches and output array SINR analogically near to the best\none.Relative to other adaptive beamforming method simulation results improve the performance.This algorithm provides\nexcellent robustness to mismatches, boosts the array system performance under non ideal situation.