Sparse Code-Domain Non-Orthogonal Random Access with Peeling Decoder

In this paper, we propose a modification of (general) sparse-coded non-orthogonal multiple access (NOMA) designs, amenable to receiver-side processing based on peeling decoding. We numerically evaluate the joint effects of sparse signature design and forward-error correction, and characterize the interplay between the system parameters such as signature sparsity, system load, and channel coding rate. The receiver processing is performed in a (turbo-like) fashion where extrinsic information is exchanged between a multi-user-detection module employing peeling decoding, and a forward error correction (FEC) module operating on the level of individual users. As the complexity of the peeling decoding procedure is lower than the general message passing algorithm (MPA) implementation based on belief propagation, it can be particularly attractive for grant-free, nonorthogonal transmissions targeting massive connectivity in the context of the Internet-of-Things.