Density-Based Clustering Resolution for Grant-Free Preamble Collision in Cell-Free mURLLC Systems.

As one of the 6 G stream services, massive ultra-reliable low-latency communications (mURLLC) can support massive short-packet data communication with high reliability and low latency, while the guarantee of its extreme demand becomes a technical difficulty. Grant-free random access (GFRA) is considered a method to reduce latency, but its performance is limited by the preamble collision caused by orthogonal resource constraints in massive communication scenarios. Due to the signal spatial sparsity of cell-free massive MIMO systems, each user being served by neighbouring access points can reduce mutual interference between collided users. This paper proposes a density-based clustering GFRA collision resolution framework in cell-free mURLLC systems, and the clusters obtained by two-dimensional information of received signal energy and geographic location are used to blind detect users for reliable access. Simulation results show that the proposed collision resolution scheme has an improved probability of successful access compared to previous resolutions, and the system is highly portable and adaptable to sporadic random traffic.