Random Spreading With Higher Order Modulation in Unsourced Random Access

Unsourced random access (URA) is crucial for future massive access scenarios. Existing URA schemes often use low-order modulation and assess performance with a limited number of active users. This work presents a pilot-free URA scheme that employs random spreading and higher-order modulation. We derive a discrete mixture generalized approximate message passing algorithm to enhance active user detection (AUD) and data detection. Simulations show that our proposed method significantly outperforms pilot-based methods in AUD, and higher-order modulation supports a large number of active users while ensuring the target per user probability of error (PUPE) performance.