Adaptive Sparse Array Reconfiguration based on Machine Learning Algorithms

The sparse array design for adaptive beamforming has been recently formulated into combinatorial antenna selection problems, which belong to notorious NP-hard problems. As the commonly deployed convex relaxation algorithms are susceptible to local optima, several trials with different initial points are conducted for the global optima. Moreover, the high computational load of optimization techniques prohibits the real-time adaptive array reconfiguration. In this work, we propose to utilize machine learning algorithms, specifically support vector machine (SVM) and artificial neural network (ANN), for solving combinatorial antenna selection problems. Numerical examples are presented to validate the effectiveness and efficiency of machine learning algorithms for sparse array design. Moreover, the SVM based antenna selection is robust against DOA estimate uncertainties.