Distributed LCMV Beamformer Design by Randomly Permuted ADMM

In recent years, distributed beamforming has attracted a lot of attention. Since each node has its own processing power, one significant advantage is the capability of distributed computing. In general, almost all distributed beamforming approaches are solving certain multi-block optimization problems. However, additional conditions are usually required to ensure convergence. In this paper, a new distributed beamforming algorithm is proposed. We first introduce the augmented Lagrangian method to implement the centralized LCMV beamformer design. Then, we propose an effective blockwise optimization method for the design of distributed LCMV beamformer based on the randomly permuted alternating direction method of multiplier (RP-ADMM). The expected convergence is obtained for distributed LCMV beamformer design without additional conditions. Numerical experiments are conducted to illustrate the performance of the proposed method.