The Non-Orthogonal Multiple Access (NOMA) technique is a promising core multiple access method for 5th Generation (5G) and beyond 5G cellular systems. In a NOMA system, it is crucial to identify pairs of users to simultaneously exploit the same frequency resource. This paper discusses an approach where the power-domain NOMA Base Station (BS) randomly selects active users and allocates transmission power based on their uplink Signal-to-Interference-plus-Noise Ratio (SINR) on the broadcast channel. Consequently, users with low SINR are defined as Cell-Edge Users (CEUs) and transmit at higher power levels compared to Cell-Center Users (CCUs). By using Monte Carlo simulations, the proposed system model with a high density of BSs in the indoor condition with walls as the main obstacles has been examined. The study demonstrates that users can achieve higher capacity under poor transmission conditions, particularly in environments with a high density and average length of walls.