Thanks to the small wavelength at millimeter

wave (mmWave) frequency, it is promising to combine massive

multiple-input and multiple-output (MIMO) with mmWave.

MmWave massive MIMO will differ the conventional

massive MIMO, due to the differences in propagation and

hardware constraints. This paper proposes a stochastic geometry

fr a mework for uating the performance in large-scale

mmWave massive MIMO networks. Based on the system model,

analytical expressions are provided for the asymptotic signal-tointerference-

plus-noise ratio (SINR) distributions in both uplink

and downlink, when the number of base station antennas goes

to infinity. Numerical results indicate a fast convergence in the

SINR distribution to its asymptotic equivalence in dense mmWave

networks. A comparison with conventional massive MIMO shows

that mmWave massive MIMO achieves a higher cell throughput

with sufficiently dense deployments.-Thanks to the small wavelength at millimeter

wave (mmWave) frequency, it is promising to combine massive

multiple-input and multiple-output (MIMO) with mmWave.

MmWave massive MIMO will differ　the conventional

massive MIMO, due to the differences in propagation and

hardware constraints. This paper proposes a stochastic geometry

fr a mework for　uating the performance in large-scale

mmWave massive MIMO networks. Based on the system model,

analytical expressions are provided for the asymptotic signal-tointerference-

plus-noise ratio (SINR) distributions in both uplink

and downlink, when the number of base station antennas goes

to infinity. Numerical results indicate a fast convergence in the

SINR distribution to its asymptotic equivalence in dense mmWave

networks. A comparison with conventional massive MIMO shows

that mmWave massive MIMO achieves a higher cell throughput

with sufficiently dense deployments.