Conflict-Aware Multi-Numerology Radio Resource Allocation for Heterogeneous Services

5G new radio (NR) introduced flexible numerology to accommodate applications with varying quality of service (QoS) requirements. However, optimizing the scheduling of services with varying delay and throughput QoS constraints remains a challenging task. Under existing proposals, supporting ultra-reliable low-latency communication (URLLC) services comes at the cost of reduced throughput offered to enhanced mobile broadband (eMBB) users. In this paper, we propose an efficient, low complexity scheduling of radio resources for URLLC when coexisting with eMBB services. We re-formulate the standard eMBB throughput maximization problem as an equivalent conflict minimization with URLLC and prove that this equivalent objective can be treated as a Bin Packing optimization problem. Moreover, in order to further increase the efficiency of resource utilization, non-orthogonal multiple access (NOMA) is also investigated for URLLC and eMBB coexistence. The superior performance of NOMA, with the superposition of services over the same resource blocks, is due to alleviating conflicts, as shown by an extensive set of numerical results.