Delay-Aware Resource Allocation for Network Slicing in Fog Radio Access Networks

To meet diverse quality of service (QoS) requirements in fifth generation (5G) wireless networks, network slicing is proposed as a cost-efficient way to allow operators to provide customized services. On the other hand, fog radio access network (F-RAN) has emerged as a promising network architecture to satisfy different QoS requirements in distinct application scenarios. To make the network slicing work more efficiently, a downlink F-RAN, equipped with a ultra-reliable and low-latency communication (URLLC) slice and an enhanced mobile broadband (eMBB) slice logically, is concerned in this paper. A delay-aware resource allocation optimization problem is formulated, aiming to maximize the throughput of the eMBB slice as well as guarantee the queuing delay performance of the URLLC slice. The delay-aware optimization problem is reformulated as a perslot joint resource block (RB) and power allocation problem, which is addressed by Lagrange dual decomposition method. Simulation results show a tradeoff between the throughput and the queuing delay of eMBB slice can be achieved while satisfying the delay constraint of URLLC slice.