Stochastic Geometry Based Performance Study on 5G Non-Orthogonal Multiple Access Scheme

To achieve a significant boost on capacity performance in the next generation (5G) cellular network, novel radio access technologies (RAT) are demanded to make the system more spectrum efficient. As a promising multiple access scheme for 5G cellular network, non-orthogonal multiple access (NOMA) has attracted extensive research attention recently. Existing works show that NOMA posses the potential to further improve system spectrum efficiency compared with the orthogonal multiple access (OMA), which is predominantly adopted by existing wireless networks. In this paper, we develop the analytical framework on system coverage and average user achievable rate in a downlink NOMA system. We explicitly consider the inter-cell interference in the study, which is a capacity limiting factor in most wireless networks but less addressed in most existing analytical work for NOMA. Additional to NOMA, the analysis on an OMA access scheme, i.e., orthogonal frequency division multiple access (OFDMA), is also conducted for comparison. Owing to the tractability of Poisson Point Process (PPP) model used in this work, all the analytical results are derived and expressed in a pseudo-closed form or a succinct closed form. The analytical results are validated by simulations and demonstrate that NOMA can bring considerable performance gain compared to OMA when success interference cancellation (SIC) error is low.