The Energy Saving Potential of Static and Adaptive Resource Provisioning in Dense Cellular Networks

Abstract

In this paper we study static and dynamic approaches to energy efficiency in dense cellular networks, where interference is one of the main limiting factors. We consider the two main approaches to energy efficiency through adaptive management of the network capacity: Base station (BS) sleeping and cell zooming. We propose an analytic framework for the assessment of the energy efficiency potential of several joint planning and management strategies. Our approach is based on stochastic geometry tools, on an approximate but accurate model of interference, and on a detailed, measurement-driven power model. For a given user density, we show how to derive the optimal BS density, and the BS transmit power which minimizes the mean power consumption of the network, while achieving a target QoS level. Through numerical evaluations, we show the potential savings enabled by joint (and disjoint) optimization of transmit power and density of active BSs. For a realistic network scenario, our approach suggests that huge energy savings are achievable by combining sleeping and zooming. In addition, we show that a static strategy, based on carefully planning the density of installed BS and their transmit power, can achieve most of the benefits of capacity tuning achievable through either sleeping or zooming. This result has a very high relevance for network operators, since it allows avoiding the feared decrease in operational lifetime which the daily switching of BS entails.