Contention-Based mMTC/URLLC Coexistence Through Coded Random Access and Massive MIMO

Radio access network slicing is considered a key feature in next-generation multiple access. In this paper, we investigate the coexistence between massive machine-type communication (mMTC) and ultra-reliable low-latency communication (URLLC) services. To meet their heterogeneous requirements, we propose a novel grant-free scheme that leverages coded random access, massive multiple-input multiple-output (MIMO) processing, and both the preamble and the power domain to enable non-orthogonal access on shared frequency and time resources. To illustrate the concept, mMTC users transmit packet replicas having different preambles in various time slots, capitalizing on the temporal domain. Meanwhile, the URLLC users apply a more aggressive strategy that leverages pilot mixture and power diversity to meet the stringent latency and reliability requirements. Contention resolution is achieved through a signal processing algorithm based on successive interference cancellation (SIC). We show that the co-design of signal processing and access protocol is crucial to meet both service requirements, and we derive fundamental limits where possible. In instances where direct derivation proves impractical, we conduct symbol-level simulations of the whole system to gain comprehensive insights. The simulations reveal that the proposed scheme can satisfy mMTC/URLLC coverage density, reliability, and latency requirements, while outperforming orthogonal allocation schemes.