A Queuing Model for Short Time Scale Sleep Mode Evaluation in 5G Base Stations

An important lever for reducing energy consumption of base stations is sleep mode, investigated in the literature since several years. In order to fully exploit the energy saving potential, advanced sleep mode that operates at a short time scale in the order of the millisecond is to be leveraged. However, short time scale sleep mode faces the issue of non negligible sleep/wakeup delays that, if not properly addressed, induce quality of service (QoS) degradation and reduce the energy gains. We develop in this paper a novel discrete time queue control scheme under different sleep mode policies and coverage scenarios, including single cell and overlapping coverage. Our numerical experiments give interesting insights about where and when to use different sleeping modes in order to achieve the largest energy gains.