ABSTRACT
There are several energy saving algorithms proposed to save energy in heterogeneous networks, but most of these algorithms achieved energy savings at the expense of service provisioning. Base stations significantly contribute to power consumption in cellular networks. This research work presented a Data Rate-Based Sleep Mode Algorithm for energy savings for a pico evolved NodeB (eNodeB) cell in aLong Term Evolution (LTE) heterogeneous network (HetNet). The algorithm was able to track traffic situations and estimate user data rate and energy consumption in the heterogeneous network. The algorithm switched the operating state of some pico eNodeB cells to sleep mode (inactive state) at low (less than or equal to 10 user equipment per macro area coverage)and medium traffic (greater than 10 user equipment per macro area coverage) during which the users are offloaded to other pico eNodeB and macro eNodeB cells to save overall energy consumption in the network. As traffic increased (greater than 10) and the average user data rate of the overall network reduced (less than 2Mbps), the pico eNodeB cells woke up (active state) to ensure service delivery was not obstructed. The work considers temporal fluctuations of traffic with a view to achieving higher energy savings. The traffic model, data rate model, and a powerconsumption modelare developed to implement the sleep mode algorithm. The developed sleep mode algorithm is implemented in Matrix Laboratory (MATLAB) R2013a. The developed algorithm was compared with the relevant specification of the 3rdGeneration Partnership Project (3GPP) Always-on scheme. The results showthat the developed sleep mode algorithm achieved an improvement of up to 75% and 50% in terms of energy savings for the pico eNodeB cells at low and medium traffic, respectively. For the overall HetNet, the improvements of 9.28% and 6.19% were achieved for low and medium traffic, respectively. The improvements of 9.45% and 6.30% are achieved in terms of the energy efficiency in 3GPP HetNet Configuration 4b for low and medium traffic while the improvement of 9.45% and 6.22% are achieved in 3GPP HetNet Configuration 1while guaranteeing a good service delivery.
TABLE OF CONTENTS
TITLE PAGE i
DECLARATION ii
CERTIFICATION iii
DEDICATION iv
ACKNOWLEDGEMENT v
ABSTRACT vi
TABLE OF CONTENTS viii
LIST OF FIGURES xi
LIST OF TABLES xiii
LIST OF APPENDICES xiv
LIST OF ABBREVIATIONS/ACRONYMS xv
CHAPTER ONE: INTRODUCTION 1
1.1 BACKGROUND 1
1.2 STATEMENT OF RESEARCH PROBLEM 5
1.3 AIM AND OBJECTIVES 6
1.4 SIGNIFICANCE OF THE RESEARCH 6
CHAPTER TWO: LITERATURE REVIEW 7
2.1 INTRODUCTION 7
ix
2.2 REVIEW OF FUNDAMENTAL CONCEPTS 7
2.2.1 LTE Network Architecture 7
2.2.2 QoS and EPS Bearers 82.2.3 Power Consumption in Base Stations 9
2.2.4Energy Saving in Base Stations 11
2.2.5Heterogeneous Networks for LTE – Advanced 12
2.2.63GPP Heterogeneous Network Topologies 14
2.2.7Base Station Sleep Mode Algorithms 16
2.3 REVIEW OF SIMILAR WORKS 18
CHAPTER THREE: MATERIALS AND METHODS 26
3.1 INTRODUCTION 26
3.2 METHODOLOGY 26
3.2.1 Traffic Model 273.2.2 Data Rate Model 28
3.2.3 Power Consumption Model 33
3.2.4Developed Sleep Mode Algorithm 37
3.3SIMULATION SETUP 41
CHAPTER FOUR: RESULTS AND DISCUSSIONS44
4.1 INTRODUCTION 44
x
4.2 VALIDATION 44
4.2.1Power Consumption 44
4.2.2 Average User Data Rate 48
4.2.3 Average Energy Efficiency 54
4.4 SUMMARY OF RESULTS 60
CHAPTER FIVE: CONCLUSION AND RECOMMENDATION 64
5.1 INTRODUCTION 64
5.2 CONCLUSION 64
5.3 SIGNIFICANT CONTRIBUTIONS 64
5.4 LIMITATIONS 65
5.5 RECOMMENDATIONS FOR FUTURE WORK 65
REFFERENCES 65
CHAPTER ONE
INTRODUCTION
1.1 BACKGROUND
During the last decade, there has been a tremendous growth in cellular networks market. The number of subscribers and the demand for cellular traffic has increased.Hence, the mobile operators find meeting these new demands in wireless cellular networks inevitable, while they save to keep their costs minimum (Hasan et al., 2011).In a wireless network, base stations (BSs) consume about two-third of total network power consumption and are logically responsible for 70% of CO2 emission from the entire network. For this reason, management of energy consumption of wireless base station has become an essential topic of discussion in the research society (Tun&Kunavut, 2014).
Figure 1.1 shows a chart of the power consumption of various components of a base station. The power supply consumes less with about 7.5%. The essence of the Air conditioning is to provide site cooling which takes up to 17.5% while signal processing
Power amplifier inc. feeder
50-80% (65%)
Air conditioning
10-25% (17.5%)
Signal Processing
(analogue+digital)
5-15% (10%)
Power Supply
5-10% (7.5%)
2
Figure 1.1:Power Consumption Distribution in Radio Base Stations (Correiaet al., 2010)
contributes about 10% of the power consumption. The power amplifier including feeder losses consumes about 65% which is highest amongst the components of the base station.
In each cell, the base station transmits common control signals and data signals to mobile users (MU), and the cell size is defined as the area in which the MUs can receive control signals from the BS. Cell size and capacity are generally static at the phase of network planning based on estimation of peak traffic load. Significant spatial and temporal fluctuations exist due to user mobility. To save energy of the whole network, the phenomenon of traffic load fluctuation implies that some base stations can be switched off when the traffic load is light (Zhinsheng et al., 2010).
Figure 1.2 shows atypical load profile across one week, which clearly exhibits significant spatial and temporal fluctuations. During either low or no traffic periods i.e., early mornings or daytime in residential areas when loads are small, many base stations are underutilized, but still consume significant amount of energy.The base stations are underutilized at night and weekends. Thus, the traffic volume of the night time is much lower than that of the daytime (Alam et al., 2012).
Figure1.2: Normalized Real Traffic Load during One Week Captured by a Cellular Operator(Oh et al., 2010)
Heterogeneous network is one of the key technologies to reduce energy consumption in a cellular network where the cell size is diversified (Xuet al., 2013).Heterogeneous Network (HetNet) could
3
mean a network comprising of different Radio Access Technologies (RATs) such as Wireless-Fidelity (WiFi), Global Systems for Mobile Communications (GSM), Universal Mobile Telecommunications System/High Speed Packet Access (UMTS/HSPA), Long Term Evolution (Advanced) (LTE/LTE-A) (Zhang, 2012). HetNets are multi-tier radio access networks in which micro and/or pico/femto cells are overlaying the macros as shown in Figure 1.3. Macro base stations are used as a baseline and provide uniform coverage of about 500m-35km and transmit power of 20W-40W (46dBm) (Zhang, 2012). Micro and pico/femto (often also referred to as small) cells are equipped with low power base stations which are deployed in hotspots to increase capacity, or in dead spots unreachable by macro base stations in order to increase coverage.The coverage area is about 40m and 20m for pico and femto cells respectively(Dini et al., 2013).
Figure1.3:HetNet Scenario with Macro Cell and Small Cells (Dini et al., 2013)
HetNets are considered as an integral part of future generation wireless networks where multiple low power, low cost small cell (e.g. femtocells, picocell) base stations (SBSs) are deployed to complement the macrocell networks. HetNets are proposed to increase the data rates and capacity to the residential areas by reducing traffic from the macrocell network (Ekti et al., 2013).
4
A picocell is a cell in a mobile phone network served by a low power (250mW(24dBm)-1W(30dBm)transmit powerand total power consumption of about 43 Watts) cellular base station that covers a small area (40m – 200m radius) with dense traffic such as a shopping mall, residential areas, a hotel or a train station. Femtocells are designed to serve much smaller areas (less than 40m) such as private homes or indoor areas (Hasan et al., 2011).
Macro BSs are deployed outdoor, normally above the rooftop level. On the other hand, small base stations are normally deployed below the rooftop level; i.e. for outdoor, they are often placed over lampposts or street furniture, and for indoor they are typically installed by the end-users in residential or enterprise settings. HetNets are more energy efficient than macro-only deployment of the same capacity, due to low power consumption of the small cells. Hence, the use of HetNets instead of macro BSs alone will achieve some energy savings (Dini et al., 2013).
Figure 1.4 shows HetNets with different deployment scenarios comprising macro-femtocell, macro-picocell, and macro-microcell with their coverage areas. The femtocells are deployed indoors.
5
[
Figure 1.4: Heterogeneous Networks (HetNets) (Hasan et al., 2011)
Small cell base stations that form an overlay layer on the existing macrocell network offer tremendous potential in terms of satisfying high data rate traffic requirements. However, small cell deployments can pose negative energy-efficiency implications if not equipped with advanced power saving mechanisms (Ashraf et al., 2010).
1.2 STATEMENT OF RESEARCH PROBLEM
There has been poor energy savings in pico eNodeB HetNets which has resulted to poor service delivery in such networks. Base station sleep mode is an indispensable part of green network technology providing significant energy savings in the overall network. Most existing sleep mode algorithms achieved energy savings without considering average user data rate and this may bring about poor service delivery. The developed algorithm will introduce a more realistic energy saving scheme through the development of a data rate-based sleep mode algorithm. The developed scheme will adopt the realisticaverage user/subscriber data rate of 2Mbps (Motorola White Paper,
6
2009) as a metric for pico eNodeB sleep mode to improve energy savings at low and medium traffic.
1.3 AIM AND OBJECTIVES
The aim of this research work is to improve energy savings in heterogeneous networks through the use ofdata rate-based pico eNodeB sleep mode.
The objectives of this research are to:
i. Adopt the traffic model of Ambrosy et al., (2012), user data rate model of Khirallahet al., (2014) and HetNet power consumption model of Richter et al., (2009).
ii. Develop abase station data rate-based sleep mode algorithm for pico cells in a HetNet using Matrix Laboratory (MATLAB) R2013a software.
iii. Validate the developed algorithm in terms of energy savings and energy efficiencythrough simulations, by comparing with thestandard 3rd Generation Partnership Project(3GPP) always-on scheme.
1.4 SIGNIFICANCE OF RESEARCH
This research work proposed a sleep mode algorithm for pico eNodeB cells in HetNet. The algorithm coulddynamically switch the operating mode of a pico eNodeB to sleep/active mode using average user data rate as a metric. The algorithm could improve energy savings and energy efficiency for low and medium traffic load when compared with the 3GPP always-on scheme in LTE HetNet respectively.This improvement is reduction in network power consumption leading to a reduction in operational cost.
7
IF YOU CAN'T FIND YOUR TOPIC, CLICK HERE TO HIRE A WRITER»
