Energy-efficient BS switching-off and cell topology management for macro/femto environments

Abstract

In the context of ever-increasing wireless data rates, energy consumption could increase fast, with severe consequences on the operator’s energy bill, in addition to environmental considerations. Energy-efficient cell breathing and BS switching-off algorithms are an important tool for energy reduction in cellular networks. In this paper, an energy-efficient cell breathing and offloading mechanism in both macrocellular and heterogeneous networks is studied. First, the need for limiting the number of Macro-BSs switched off and the control of cell sizes are analyzed. An approach based on the use of a traffic threshold from a technique previously proposed by our team is applied in order to control the BS switching-off aggressiveness. One reason for limiting this aggressiveness is the trade-off we point out between the existing RAN power consumption and the transmission power in uplink for mobile devices. Then, in the context of macro/femto deployments, the BS switching-off possibilities are extended by means of the extra capacity from an added femtolayer under regulated control access. We explore the impact of access policies from 3GPP CSGs (Closed Subscriber Groups) on the network performance and how CSG features may be used by operators to set aspects such as pricing policies and QoS provision levels. In this article, our goal is to provide a combined approach of both techniques in a single framework for macro/femto environments. The obtained results are analyzed showing the importance of proper tuning of energy-efficient algorithms in order to guarantee convenient energy savings and maintaining good QoS levels.

Introduction

Nowadays, it is very well known how the dynamics of growth in the mobile sector are under significant pressure from an increase in the volume of traffic. According to [1] the mobile data traffic increased 81% from 2012 to 2013, which represents around 18-fold times size of the Internet in early 2000s. As the traffic increases, a more significant investment becomes necessary from the operator side to deploy larger radio access networks, which also comes with an associated increase in the operational expenditure (OPEX). An important cause of this growth of costs is the price of energy to run such an infrastructure, in most cases supplied by fossil fuel-derived sources, which brings an ecological impact due to its associated CO₂ emissions. All these issues have become an important preoccupation for governments and regulation bodies that have brought about initiatives like in the case of the Code of Conduct on Energy Consumption of Broadband Equipment promulgated by European Commission [2]. In addition, it comes along with a technical necessity to provide standards and reference frameworks like the ETSI/3GPP TS 32.551 [3], which provides concepts and requirements for Energy Saving Management (ESM) or other documents like the 3GPP TR 32.834 [4], which goes deeper on the energy savings and the search for balance in scenarios with a diversity of Radio Access Technologies (RAT). These initiatives have brought about a strong interest from the different players in the ICT sector to join forces and bring technological solutions, creating a new field of engineering where computer science, electronics and telecommunications expertise converge: the Green Radio. Many research teams and consortiums worldwide have proposed solutions with a diversity of approaches providing a very rich literature on the topic [5]. Those approaches start with enhancements to electronic components and internal device architectures, where significant effort is made on the power amplifier. Above that level, the optimization of transmission processes and energy-aware radio resource allocation is placed. Finally, at the top we have the infrastructure and topology adaptation that consists of customizing the radio access network by using approaches such as BS switching-off, cell breathing and the use of heterogeneous networks composed of macro/small-cell devices, relays or renewable energy supplied equipment.

For the specific case of cell breathing, a significant work exists which proposes energy-efficient switching-off schemes supported by this technique. Historically, cell breathing is an approach that has existed since the 90s [6], [7], originally conceived to reduce the global interference by means of load balancing and power adjustment in the cellular deployment. In recent years, it has been proposed as an energy-efficient approach (see for instance [8], [9], [10]). Here, the cell sizes and the number of active BSs are adapted to fluctuations in traffic throughout the coverage area, whereas some unnecessary BSs with zero or very low loads are deactivated by means of redistributing the traffic toward BS neighbors. These BS neighbors must remain active, increasing their cell sizes in order to guarantee coverage.

The purpose of this paper is to explore in depth different strategies that make it possible to regulate or increase the number of switched-off Macro-BSs when an energy-efficient cell-breathing approach is applied to a radio access network. Firstly, one of our objectives is to study the existing trade-off between the number of switched-off BSs and the levels of transmission power when cell breathing is applied in a mobile network. In this first case, our intention is to limit the number of deactivated Macro-BSs in order to have control over the cell sizes of the remaining active Macro-BSs, which has a direct relationship with the transmission distances between the Mobile Station (MS) and the Macro-BS. For this first part of our work, we analyze these aspects, taking as a point of reference a technique called Mobile-Aware Distributed BS-based Cell Breathing (MA-DBCB). This mechanism, previously proposed by our team in [11], has among its different features the possibility of blocking the deactivation of any BS by the use of a load threshold. Given the distributed characteristics of our algorithm, this threshold can be set globally and locally for each BS in the network. Our focus is the analysis of this strategy on uplink where it has a significant impact on MS transmission power levels and brings associated consequences for device performance (e.g. battery lifetime, electromagnetic compliance).

Later on, a second proposal by our team called femto-DBCB is analyzed [12]. Here, our goal is to increase the number of switched-off Macro-BSs with reduced impact on an already defined cell-breathing scheme. This is done by adding an associated deployment of small BSs, i.e. femtocells, and then combining cell-breathing at the macrocellular layer with a second means of traffic redistribution based on macro-to-femto offloading. Then, the inclusion of the 3GPP Closed Subscriber Groups concept [13] is deeply studied, with the purpose of analyzing how blocking due to access policies can impact the performance of the proposed strategy. When CSG policies are in place, the full access to the femtolayer becomes only possible for allowed users with privileged access to radio resources, whether because of being owner of a home installed femtocell or for having paid for a preferential service. On the contrary, any other regular user counts with a restricted access to some public released resources from the femtolayer, that in the case of not being available the only option left is to rely on the resources from the macrocelullar layer. It is pointed out how blocking policies become an additional knob to set the network functioning and in that sense, operators must define their access and pricing policies accordingly. Finally, the two proposals are put together to provide a combined framework for heterogeneous networks and a deep analysis is provided throughout simulations.

The outline of this paper is as follows. In Section 2 we justify the need for strategies to limit or increase the number of switched-off Macro-BSs in the context of energy efficiency. In Section 3 details of the strategies proposed for making it possible and the integrated framework is provided. In Section 4, different simulated scenarios are provided for each of the strategies followed by the integrated approach analysis and the results found. Finally, conclusions are provided in Section 5.