Power-Tolerant NOMA Using Data-Aware Adaptive Power Assignment for IoT Systems

Non-orthogonal multiple access (NOMA) is a promising candidate for future wireless networks due to its ability to improve the spectral-efficiency and network connectivity. Nevertheless, the error rate performance of NOMA depends significantly on the power assignment for each user, which requires accurate knowledge of the channel state information (CSI) at the transmitter, which can be challenging for several applications such as wireless sensor networks (WSNs) and Internet of Things (IoT). Therefore, this paper proposes a power-tolerant NOMA by adaptively changing the signal power of each user to reduce the system sensitivity to inaccurate power assignment. The power adaptation in the power-adaptive NOMA (PANOMA) is performed based on the transmitted data, and it does not require accurate CSI. To quantify its potential, the bit error rate (BER) and the lower bound capacity performance, over Rayleigh fading channels, are derived in exact closed-forms for two and three users scenarios. The results demonstrate that PANOMA provides a tangible BER performance improvement over conventional power-domain NOMA when both schemes use sub-optimal power assignment, which is typically experienced in practical scenarios involving channel time variation and CSI estimation errors. Specifically, it will be shown that both schemes provide similar BERs using optimal assignment, but the PANOMA offers BER reeducation by a factor of 10 for certain scenarios when sub-optimal power values are assigned. The integrity of the analytical results is verified via matching extensive Monte Carlo simulation experiments.