A spatio-temporal sleep mode approach to improve energy efficiency in small cell DenseNets

Data traffic has been increasing exponentially and operators have to upgrade their networks to meet the prevailing demand. This effort entails deploying more base stations (BSs) to meet the increasing traffic. The resulting capital expenditures (CAPEX) and operational expenditures (OPEX) have limited operator revenues. In addition to the energy costs, the associated greenhouse gas emissions have raised environmental concerns. In this paper, we use system-level simulations to investigate different sleep mode mechanisms that can address both capacity and energy efficiency (EE) objectives in dense small cell networks (DenseNets). These sleep mode approaches are applied to a long-term traffic profile that is obtained from real world network traffic. We then design a mechanism that determines the required BS density in response to the variable long-term traffic profile. Our results reveal that significant energy savings are possible when sleep mode mechanisms are applied based on the prevailing traffic in both time and space domains.