A Low-Complexity Interference Cancellation Approach for NOMA

Non-orthogonal multiple access (NOMA) has been studied during the standardization of the new radio (NR), fifth generation (5G) of mobile communications systems. It has been shown that NOMA could potentially offer a robust performance in highly interfered scenarios. Despite the observed capacity gain and performance, the issue of NOMA receiver complexity may remain as a potential impediment to adoption of NOMA. The proposed advanced receivers for NOMA are more complex than regular interference cancellation receivers used in base stations. A typical advanced receiver for NOMA is composed of a core multi user detector (MUD) and a forward error correction (FEC) decoder that iteratively attempts to estimate and suppress multi-user interference. At each iteration, interference from previously decoded users are reconstructed and cancelled allowing remaining users to be successfully decoded. Among the proposed advanced receivers for NOMA, receivers using expectation propagation algorithm (EPA) has attracted much attention due its high performance. However, high complexity of such iterative decoding architectures has shadowed their performance and capability. In this paper, we investigate the performance of a simplified version of such iterative MUD receiver such as EPA receiver. According to our evaluations, the simplified receiver can maintain a similar performance as the original EPA receiver for packets with low code rates in both low and high interference environment. In case of packets with a higher code rate, the performance is only marginally degraded.