In Non-Orthogonal Multiple Acess (NOMA), power allocation and user pairing is the major challenge. We have addressed this issue in the proposed work. We have formulated a power optimization problem for the sum-rate maximization with power and minimum rate constraints and solved it using a Genetic Algorithm (GA). The results of the proposed work indicate that a better fairness index can be achieved in the investigated work compared to existing solutions. Along with this, we have developed an algorithm for user pairing based on the Hungarian work assignment problem. It is emphasized that better pairing in the NOMA scheme achieve a higher sum rate. In the underlined work we have extended the Hungarian work assignment problem for 3 user clustering cases. It is observed that obtained results of the proposed work in terms of total sum rate outperform the Near- Far (NF) and Near-Near (NN) user case for various channel conditions and the number of users. A proposed Optimum Pairing Scheme (OPS), performs well for 2 user pairing as well as 3 users clustering to achieve better capacity along with a good fairness index