Massive MIMO Systems for 5G Communications with Optimal Spectrum Efficiency

摘要

Massive Multi-input and Multi-output(MIMO) is an assuring technique to enlarge spectral-efficiency (SE) for wireless networks, by assembling antenna arrays using more active elements on base stations then by performing the coherent-transceiver processing on each base station. A common criterion exists that the systems are supposed to contain an order of magnitude of more M antennas than the K scheduled users (SUs), since channels of the user's are expected to be close-orthogonal when M / K > 10, but this criterion which optimizes the SE has not been proved. Here, we investigate how the optimal-number of K* SUs depends on M antennas and new SE expressions are also derived to facilitate effective system-level scrutiny through power-control, user positions and random pilot-reuse. Use of K* in the large-M system antennas is obtained in closed-form. Concurrently the simulations are also used to illustrate what occurs at finite M, in dissimilar interference conditions, with dissimilar processing techniques and dissimilar pilot-reuse factors, and processing patterns. Up to great portion of the coherence block have to be committed to pilots as well as the optimal M/K are limited to 10 in various circumstances of practical purpose. Interestingly, K* relies effectively on the processing technique and therefore it is unreasonable to compare distinct techniques using the same number of K.