Blind Digital Beamforming Techniques for Next Generation Communication Systems

Beamforming has become an essential technique for next-generation wireless systems. Digital beamforming is known to provide optimal beamforming performance and is robust with advanced signal processing techniques. In this paper, we focus on developing various algorithms which blindly estimate the symbols. Blind beamforming algorithms proposed for a single user do not put any constraint on the structure of the signal, unlike already existing source separation methods. For multiple users, a threshold-based Iterative Least Squares with Projection (ILSP) is proposed, an advancement to ILSP where the symbols near the decision boundary are not wrongly projected. Furthermore, when the users are located closely in the spatial domain, we present an algorithm which utilises Estimation of Signal Parameters via Rotation Invariance Technique (ESPRIT), a subspace-based method to estimate the Direction of Arrival (DoA) precisely. Consequently, the symbols are estimated using the interference cancellation technique. Finally, we develop a Neural Networks (NN) based beamforming algorithm that predicts the symbols of the user of interest.