Low-complexity variable forgetting factor mechanisms for adaptive linearly constrained minimum variance beamforming algorithms

摘要

In this work, the authors propose two low-complexity variable forgetting factor (VFF) mechanisms for recursive least squares-based adaptive beamforming algorithms. The proposed algorithms are designed according to the linearly constrained minimum variance (LCMV) criterion and operate in the generalised sidelobe canceller structure. To obtain a better performance of convergence and tracking, the proposed VFF mechanisms adjust the forgetting factor by employing updated components related to the time-averaged LCMV cost function. They carry out the analyses of the proposed algorithms in terms of the computational complexity and the convergence properties and derive an analytical expression of the steady-state mean-square-error. Simulation results in non-stationary environments are presented, showing that the adaptive beamforming algorithms with the proposed VFF mechanisms outperform the existing methods at a significantly reduced complexity.