PROJECT TITLE :

Adaptive reweighting homotopy algorithms applied to beamforming

ABSTRACT:

We tend to develop adaptive beamforming algorithms that are robust against sensor failure and unwell-conditioning within the autocorrelation matrix (common in low-rank interference eventualities). Each goals are achieved simultaneously through the utilization of l1 regularization. The algorithms are primarily based on the complex adaptive reweighting homotopy technique. We also develop iterative versions of the algorithms that use properties of homotopy l1 solvers and dichotomous coordinate iterations to scale back significantly the computational complexity, compared with different regularization strategies.