Edge Users Spectrum Efficiency Maximization Using Spatially Aligned Carrier Aggregated NOMA in Dense Multi-Cell Networks

This article studies a novel spatially aligned carrier aggregated non-orthogonal multiple access (SACA-NOMA) scheme for spectrum distribution and fair power allocation among grouped users. The proposed SACA-NOMA scheme integrates NOMA, carrier aggregation (CA) and joint transmission coordinated multipoint (JT-CoMP). It uses spatial differentiation between multiple NOMA pairs to combat interference. The main objective of the scheme is to maximize the spectral efficiency and quality of service (QoS) of cell-edge users by grouping them with multiple NOMA pairs. NOMA on top of CA assisted JT-CoMP is a promising contender for next-generation cellular networks because it can address the issues of spectrum scarcity and edge-user QoS. Two challenges arise from this integration when NOMA pairs are grouped with cell-edge user. Among these include an increase in interference at cell-edge user and complexity of implementing successive interference cancellation. These challenges can be addressed by exploiting the additional spatial domain. Base stations can cooperate to achieve interference alignment at cell-edge users through zero-forcing pre-coding. Similarly, a grouping scheme is presented wherein number of users in NOMA based pairing never exceeds two on any spatial axis. To this end, we formulated a joint problem of spectrum allocation, power control and BS's admission control into proposed scheme, with an aim to enhance the spectral efficiency of edge users. A fair power allocation and admission control algorithm is developed to solve the mixed integer non-linear fractional problem. The obtained numerical results show that the proposed SACA-NOMA scheme offers the cell-edge users a significant performance gain of about 60% and 300% when compared to the multi-cell hybrid NOMA-OMA and single cell OMA schemes, respectively. The proposed scheme is also validated for a typical macro-cell in an urban environment.