Robustness and Performance Enhancement in Smart Antenna Systems through Novel Beamforming Techniques

Abstract

Beamforming is a technique used to process signals received by an array of antennas. It amplifies signals coming from desired directions while reducing the strength of signals coming from undesired directions. The algorithms used to perform realtime beamforming are called adaptive beamformers. The adaptive beamforming algorithms are sensitive to certain circumstances, such as the steering vector errors, small samples for parameter estimation, non-stationary interference signals, increase sidelobe level (SLL), multiple desired signals, etc. The steering vector errors caused by multiple factors such as mismatch in the direction of arrival (DOA) of signal of interest (SOI), imperfect antenna array calibration, perturbation in propagation medium, sensor position uncertainty, etc. Therefore, to overcome these problems, robust adaptive beamformers are required. In this work, various robust beamformers have been developed to improve the performance of beamforming algorithms in different scenarios. The foremost cause of the steering vector error is DOA mismatch, so the actual direction of the desired signal is estimated iteratively through proximal gradient approach. The regularization function is utilized to modify a hyper-parameter for each iteration and also guarantees that the estimated direction is sufficiently close to the actual direction in case of large DOA mismatch. In a moving interference environment, traditional algorithms become ineffective as the interferences may shift outside the null created by the beamforming algorithm. Expanding the nulls effectively resolves the issue of moving interferences. The process of null widening involves modifying the steering vector by introducing a taper matrix of imaginary interferences