Non-Orthogonal Multiple Access (NOMA) and massive Multiple Input Multiple Output (mMIMO) have been considered or addressed as a promising candidate for the forthcoming generation networks speci cally fth generation (5G) networks and beyond 5G. In this thesis, we have discussed and analyzed the performance of downlink NOMA for massive MIMO Systems in contrast to conventional MIMO systems. Furthermore, we have also explored the performance gains of MIMO NOMA compared to its traditional MIMO OMA counterpart. Moreover, we considered the performance gains of downlink NOMA over conventional OMA schemes particularly OFDMA systems. Since the early contributions of this area mainly focused on the scenario of Single Input Single Output (SISO) and conventional multi-user MIMO systems, it is of particular interest to conduct the performance analysis of NOMA for massive MIMO systems by comparison with conventional MIMO systems. Nevertheless, we have discussed the performance of downlink NOMA for mMIMO in contrast to traditional multiuser MIMO with NOMA by considering analytical model. Users are grouped into di erent clusters by combining any two users with same spatial direction or having higher correlation between their channels but having di erent channel conditions. In addition, we have investigated the performance of two user per cluster in downlink NOMA with a base station equipped with large number of antennas. We have used 64 number of antennas at the BS in the case of massive MIMO NOMA, but in the case of multiuser MIMO NOMA, we used three di erent con gurations of antennas at the BS, i.e., three number of antennas at the BS and at the user side. Perfect channel state information is assumed in both massive MIMO NOMA and multiuser MIMO NOMA, and we have allocated less power for the near users where we have allocated more power for the weak users. System capacity or ergodic sum capacity is the major points that we have considered in order to analyze the performance of downlink NOMA for massive MIMO systems in contrast to conventional MIMO systems. The numerical conclusions reveal that NOMA outperforms OMA in both single cell and multi-cell environments, and multiuser MIMO NOMA outperforms multiuser MIMO OMA. Moreover, we have also noticed that massive MIMO NOMA outperforms MIMO NOMA even when the number of antennas at the BS and at the user side in the case of MIMO NOMA is about twelve.