Impact of spatio-temporal power sharing policies on cellular network greening

Abstract

In this paper, we analyze short-time scale BS (Base Station) power saving controlling transmit power and scheduling users of each BS on multi-cell multi-carrier cellular networks. Although the transmit power of a typical macro BS is lower (10W-20W) than total BS power consumption (500W- 2000W), the transmit power also exerts influence linearly on the power consumption of power amplifiers and cooling systems and battery backup and power supply loss in the BS. Even with a transmit power reduction, users can expect high performance when they are in the interference-limited region and appropriate power control and user scheduling algorithms are adopted. Through simple simulations under EQ (equal power allocation) and IM (interference management) cases, we verify that more than 76% throughput can be obtained for edge and center users only with a 1/5 of transmit power. Furthermore, we propose four BS power sharing policies which are novel greening policies exploiting temporal and spatial degree of freedom with fixed fairness criterion (proportional fair) and present power allocation and user scheduling algorithms to meet asymptotically near optimal solutions of given framework. This study is of great importance in that the pressure on the CO2 emission limit per nation increases, e.g., by Kyoto protocol, which will ultimately affect the power budget of a wireless service provider. Through extensive simulation, we verify that the impact of proposed power sharing policies on BS power saving. Besides, we show more outstanding greening performance of IM and spatio-temporal power sharing than IM and no spatio-temporal power sharing through the simulations under the part of the Manchester city, UK (United Kingdom) irregular BS deployment and power consumption scenarios. Furthermore, we verify greening performance of our frameworks on the small cell (such as micro and femto cell) which is an inevitable trend in the next generation cellular network to maximally exploit the spectral resources. A promising conclusion in drawn that higher GAT and greening efficiency are achieved with smaller cell sizes when applying IM and spatio-temporal power sharing with network-level average power constraint than EQ with no spatio-temporal power sharing.