A Quadratically Constrained Stochastic Gradient Algorithm for Beamforming in Mobile Communications
Ciro Pitz, Eduardo Luiz Ortiz Batista, Rui Seara · IEEE Transactions on Circuits & Systems II Express Briefs · 2017
This brief presents a new adaptive beamforming algorithm for mobile communication systems with antenna arrays. Such an algorithm belongs to the constrained stochastic gradient class of algorithms, in which the maximization of the signal-to-interference-plus-noise ratio (SINR) is carried out by using stochastic gradient optimization strategies along with instantaneous cost functions related to the SINR. The main novelty of the proposed algorithm is the use of an adaptive quadratic constraint that allows obtaining enhanced solutions for both transient and steady-state phases of the iterative process. As a consequence, the proposed algorithm, termed adaptive-projection quadratically constrained stochastic gradient (AP-QCSG) algorithm, is capable of outperforming other constrained stochastic gradient (CSG) algorithms from the literature. Simulation results are presented aiming to confirm the effectiveness of the proposed approach.