Resource Allocation for eMBB/URLLC Coexistence in Massive MIMO Industrial Automation

Enhanced mobile broadband (eMBB) and ultrareliable low-latency communications (URLLCs) are two critical service types in industrial automation. In a closed-loop control system, device-to-device (D2D) communication is typically employed for direct transmission due to its low-latency requirements. However, this approach does not leverage the large-scale antenna gains of massive MIMO cellular systems. To address this limitation, we introduce a multiconnectivity network that integrates both cellular and D2D links to serve URLLC sensors while accommodating the transmission needs of general eMBB traffic. Since the D2D link serves as the primary link for URLLC transmission in a multiconnectivity setup, we first analyze the packet loss probability components for single-D2D URLLC link. Then, we formulate an optimization problem to maximize the sum channel capacity of eMBB sensors while satisfying URLLC QoS requirements. A suboptimal power and spectrum allocation scheme is proposed to solve this coexistence problem of single-D2D URLLC and cellular eMBB transmission. For multiconnectivity, we examine the packet loss probability and present two transmission frameworks based on selection combining (SC) and maximal ratio combining (MRC). Simulation results validate the properties of the optimal solution for the relaxed problem and demonstrate the performance gains of multiconnectivity over single-D2D links.