Network slicing in aerial base station (UAV-BS) towards coexistence of heterogeneous 5G services

Abstract

Unmanned aerial vehicle base stations (UAV-BSs) empowered with network slicing capabilities are presented in this work to support three heterogeneous classes of 5G slice service types, namely enhanced mobile broadband (eMBB), ultra-reliable low-latency communication (uRLLC), massive machine-type communication (mMTC). The coexistence of eMBB, uRLLC and mMTC services multiplexed over common UAV-BS radio resources leads to an incredibly challenging downlink scheduling problem due to the underlying trade-off of end-user requirements in terms of coverage, traffic demand, data rates, latency, reliability, and UAV-specific constraints. To this end, a modular and customizable two-phase resource slicing optimization framework is proposed for UAV-BS known as gEneral rAn Slicing optImizEr fRamework (EASIER) decomposed into: (i) resource optimizer (RO) and (ii) scheduling validator (SV). The reciprocation of RO and SV guided by above split optimization model can generate efficient scheduling decisions that benefit constrained UAV platforms in terms of finite computation and endurance. Furthermore, prioritizing per slice user acceptance rate, our results show that EASIER not only adheres to slice-specific SLAs (service level agreements) specified by the slice owners (i.e., tenants), but also benefit from efficient UAV-BS positioning to improvise service offering by 15% as compared to a slice-agnostic “default” positioning.