Low Complexity Signal Detection for Massive-MIMO Systems

Maximum likelihood detection is infeasible in uplink multiuser massive multiple-input and multiple-output (m-MIMO) systems due to the large dimension of the MIMO systems. Accordingly, suboptimal or near-optimal alternatives like linear minimum mean square error (LMMSE) detector and Zero Forcing (ZF) are used. However, the LMMSE and the ZF detectors need matrix inversion, which is computationally costly. We propose two detection schemes for massive MIMO, which compute an approximate inverse based on the Cayley-Hamilton theorem, and have quadratic complexity in the number of users. Simulation results exhibit the similarity of the BER performance of the proposed schemes to that of the ideal ZF or LMMSE.