A Novel Power Allocation Strategy for Finite Alphabet in MIMO Systems

One of the significant benefits of multiple-input-multiple-output (MIMO) for wireless applications is the high spectral efficiency it can achieve. Water filling (WF) has long been deemed as the optimal power allocation strategy to maximize the spectral efficiency. However, the high complexity of this algorithm is annoying in the case of finite alphabet. The current paper suggests a novel power allocation strategy in the context of finite alphabet that is much simpler than WF. It can achieve 95% of the spectral efficiency of WF in the i.i.d. Rayleigh fading channel, and even outperform WF in highly correlated fading channel, e.g. the line-of-sight (LOS) Ricean fading channel