Analysis of Fair-NOMA Power Allocation: a Comparative Study Between NOMA and OMA Systems

Abstract

Non-Orthogonal Multiple Access (NOMA) system has been very interesting technique recommended for the fifth generation (5G) of wireless mobile communication networks. The strategy of a suitable multiple access technique is one of the most important aspects in improving the system capacity. Multiple access techniques can broadly be categorized into two different approaches, namely: OMA and NOMA. In this paper a downlink communication system composed of a base station and M users where each one is equipped with a single antenna is considered, and multiple users system are characterized by multipath communication known as Rayleigh fading channel which can be described by a complex fading channel coefficient. It is well known that NOMA fairness feature implies that, with NOMA approach, near and far mobile users from a base station are able to achieve at least the possible information capacity similarly to OMA users. Fair-NOMA is described by fairness index (F) and a power allocation coefficient (a) limiting the fairness region R bounded by extreme values amin and amax, this fairness index is given as a function of NOMA power allocation coefficient, SNR ratio and channel gains of NOMA users. Maximum fairness index is obtained for a power allocation coefficient belonging to the fairness region, which is large for weak SNR ratios, and slight for high levels. In addition, the calculation of NOMA and OMA ergodic capacities at fairness bounds, indicates that NOMA approach always outperforms OMA one. Further, fairness index evolution is considered in function of channel gains, power allocation coefficients and SNR ratio levels, giving as result, optimal fairness can be reached by a judicious power allocation to the furthest user from the base station.