Massive MIMO has become a core technology for the next generation of wireless communications, and the non-linear group decoding schemes can achieve better spectral and energy efficiency, which in turn leads to higher data rates, especially with the assistance of non-orthogonal frequency multiple access (NOMA). The purpose of this paper is to study the problems of formulating performance of such systems. In particular, this paper, closely approximates the achievable rate of large-scale MIMO systems occupied by clients (users) using the optimal successive group decoding (OSGD) scheme to mitigate the co-channel interference. In addition, we produce a simple mathematical representation to accurately approximate the outage probability of this system. We also formulate the effective capacity of the considered system in a simple way. The results obtained by the Monte Carlo simulation show the validity of the derived formulas. The findings of this paper show that increasing the number of antennas linearly increases the expected value of the achievable rate and decreases its variance. In addition, the systems capacity is increasing in a logarithmic manner with the allocated power. Also, we notice that the distribution of the random achievable rate tends to be closely approximated with log-normal distribution as the number of antennas increases.