Low-complexity Collision and Data Detection for Contention-based Grant-free Access in Massive IoT

Grant-free non-orthogonal multiple access (GF-NOMA) is a promising solution for supporting Internet of Things (IoT), addressing the challenges of sporadic device activity and limited channel resources. This work focuses on contention-based GF-NOMA, where active devices randomly select resources from a non-orthogonal pool and transmit without a grant request. This can lead to collisions among devices that randomly choose the same channel resource. To this end, a scalable, low-complexity method for collision detection and data decoding in power-domain GF-NOMA is proposed by employing intelligently located pilots in the GF-NOMA resource unit. Detailed simulation results demonstrate the performance gains of the proposed approach.