Robust Precoder Design for User-Centric Network Massive MIMO With Symplectic Optimization

In this letter, we propose a symplectic optimization approach for robust precoder design in the user-centric network (UCN) massive MIMO (mMIMO) system. The system implementation of the UCN mMIMO is simplified and the dimension of the precoder is reduced, since each user terminal (UT) communicates with a properly selected subset of all base stations (BSs). We use an a posteriori channel model that incorporates channel aging to enhance system robustness. Based on this, we develop a robust precoder design method with symplectic optimization, which addresses the high complexity of the matrix inversion in traditional linear precoders optimization. Moreover, we use the relativistic dynamical system in symplectic optimization to get a more stable algorithm. We iteratively obtain the precoder vectors directly by utilizing the transformation between potential and kinetic energy in the dynamical system. Simulation results indicate that the proposed approach significantly outperforms the WMMSE precoder in mobility scenarios with lower computational complexity.