EESO: Energy Efficient System-resource Optimization of Multi-Sub-Slice-Connected User in 5G RAN

Abstract

5G has defined three broad service categories, eMBB, mMTC and URLLC. The network functions and system resources for a service category between user equipment (UEs) and 5G base stations (gNodeB) defines a network slice. Each UE depending on its Service Level Agreement (SLA) logically uses portion of system resources allocated per network slice to form sub-slices. To support a diverse range of applications, we propose that the UE can simultaneously be connected and grouped to multiple sub-slices per network slice (multi-sub-slice-connected UE). On the contrary, maintaining the above multi-sub-slices services connection for a single UE leads to increased energy consumption. To the best of our knowledge, existing literatures have not proposed an extensive optimization model to achieve energy efficient system resource optimization for multi-sub-slice-connected UE across varying channel characteristics. In this paper, we propose an Energy Efficient System-resource Optimization (EESO) algorithm. From novel EESO algorithm, we derive closed-form models for required throughput and energy efficiency in terms of optimum resources per multi-sub-slice-connected UE service. Extensive simulations based on closed-form models, reveal that EESO ensures minimum and maximum performance gains of 3.8 and 10.1 times, with respect to, average cell throughput and energy efficiency compared to well-known latest algorithms in literature.