A Separation Architecture for Achieving Energy-Efficient Cellular Networking

Abstract

Energy-efficient cellular networking has received considerable attention recently in hope of finding novel solutions to reduce network energy consumption. In this paper, a case study is conducted for a separation architecture in which two types of base stations (BSs) simultaneously serve a geographic area, one for providing reliable coverage and the other for handling user traffic. Based on a postulated BS power model, we demonstrate that the separation architecture, when replacing the conventional macro BS with a light-weight coverage BS (CBS) and multiple traffic BSs (TBSs), significantly reduces the overall energy consumption of a cellular network. Numerical results suggest that the separation architecture can usually reduce the energy consumption by 50% or even more compared with conventional macro BS. We then investigate dynamic TBS adaptation (i.e., BS switching on/off), based on traffic load fluctuations. Closed-form results are derived to suggest approximately linear adaptation of the intensity of TBSs in the separation architecture. Moreover, we consider the optimal deployment of TBSs over a long time scale, and derive closed-form results for the optimal intensity of TBSs for a given user intensity. Extensive simulations demonstrate that the proposed separation architecture is a promising solution to enable energy-efficient cellular networking.