ETSI TR 103 117 V1.1.1 (2012-11)

Environmental Engineering (EE); Principles for Mobile Network level energy efficiency

DTR/EE-EEPS004

General Information

• Status: Published
• Publication Date: 07-Nov-2012
• Technical Committee: EE EEPS - EE Eco Environmental Product Standards
• Current Stage: 12 - Completion
• Due Date: 07-Nov-2012
• Completion Date: 08-Nov-2012
• Mandate: M/462
• Mandate: M/495

Standards Content (Sample)

Technical Report
Environmental Engineering (EE);
Principles for Mobile Network level energy efficiency

2 ETSI TR 103 117 V1.1.1 (2012-11)

Reference
DTR/EE-EEPS004
Keywords
Energy Efficiency, radio, access, WCDMA, LTE
ETSI
650 Route des Lucioles
F-06921 Sophia Antipolis Cedex - FRANCE

Tel.: [phone]  Fax: [phone]

Siret N° 348 623 562 00017 - NAF 742 C
Association à but non lucratif enregistrée à la
Sous-Préfecture de Grasse (06) N° 7803/88

Important notice
Individual copies of the present document can be downloaded from:
http://www.etsi.org
The present document may be made available in more than one electronic version or in print. In any case of existing or
perceived difference in contents between such versions, the reference version is the Portable Document Format (PDF).
In case of dispute, the reference shall be the printing on ETSI printers of the PDF version kept on a specific network drive
within ETSI Secretariat.
Users of the present document should be aware that the document may be subject to revision or change of status.
Information on the current status of this and other ETSI documents is available at
http://portal.etsi.org/tb/status/status.asp
If you find errors in the present document, please send your comment to one of the following services:
http://portal.etsi.org/chaircor/ETSI_support.asp
Copyright Notification
No part may be reproduced except as authorized by written permission.
The copyright and the foregoing restriction extend to reproduction in all media.

© European Telecommunications Standards Institute 2012.
All rights reserved.
TM TM TM
DECT , PLUGTESTS , UMTS and the ETSI logo are Trade Marks of ETSI registered for the benefit of its Members.
TM
3GPP and LTE™ are Trade Marks of ETSI registered for the benefit of its Members and
of the 3GPP Organizational Partners.
GSM® and the GSM logo are Trade Marks registered and owned by the GSM Association.
ETSI
3 ETSI TR 103 117 V1.1.1 (2012-11)
Contents
Intellectual Property Rights . 5
Foreword . 5
Introduction . 5
1 Scope . 7
2 References . 7
2.1 Normative references . 7
2.2 Informative references . 7
3 Definitions and abbreviations . 9
3.1 Definitions . 9
3.2 Abbreviations . 9
4 Definition of the network in the scope . 11
5 Metrics and data for network level energy efficiency . 12
5.1 Metrics and energy efficiency in literature . 12
5.2 Proposal of metrics for network level energy efficiency . 14
6 Network level measurements in real networks . 15
6.1 Measurement procedure . 15
6.2 Definition of the area of the measurement . 16
6.2.1 Area of measurement used in simulations . 16
6.2.2 Partial area measurement definition proposal . 17
6.3 Measurement set-up alternatives for QoS KPIs. 17
6.3.1 Measurement with network internal mechanisms . 18
6.3.2 Measurement from probes . 19
6.3.3 Measurement from terminal reporting . 20
6.3.4 Measurement by end user evaluation . 21
6.3.5 Measurement by dedicated test user evaluation . 21
6.3.6 Measurement by combining system tools and network UEs . 22
6.4 Power consumption measurements . 22
6.5 EE KPI calculations from measurements . 22
6.5.1 Case of QoS measurements with network internal mechanisms . 22
6.5.2 Other cases . 23
6.6 Recommended measurement setup . 23
7 Lab network measurements . 24
7.1 Benefits and drawbacks of lab network measurements . 24
7.2 Alternative solution: live sub-network measurements . 25
7.3 Summary . 26
8 Extension to global scale . 27
8.1 Scenario definition . 27
8.2 Ratio of different scenarios in an example network . 28
8.3 Example of extrapolation . 29
9 Recommendations for future work . 30
Annex A: OPERA-Net . 32
A.1 The OPERA-net project . 32
A.1.1 The OPERA-Net base station power model . 35
A.1.1.1 Description of macro base station sites . 35
A.1.1.2 Site support system . 35
A.1.1.3 Radio Base Station . 36
A.1.1.4 The BTS power consumption model . 37
A.1.1.5 BTS model variation for independent sector load . 38
A.1.1.6 Ideal BTS model . 38
ETSI
4 ETSI TR 103 117 V1.1.1 (2012-11)
Annex B: Earth project . 39
B.1 EARTH Major Results . 39
B.1.1 Description of the model . 40
B.1.2 Rationale. 42
B.1.3 Domain of validity. 43
B.1.4 Use case examples . 43
Annex C: MDT to estimate QoS . 44
History . 46

ETSI
5 ETSI TR 103 117 V1.1.1 (2012-11)
Intellectual Property Rights
IPRs essential or potentially essential to the present document may have been declared to ETSI. The information
pertaining to these essential IPRs, if any, is publicly available for ETSI members and non-members, and can be found
in ETSI SR 000 314: "Intellectual Property Rights (IPRs); Essential, or potentially Essential, IPRs notified to ETSI in
respect of ETSI standards", which is available from the ETSI Secretariat. Latest updates are available on the ETSI Web
server (http://ipr.etsi.org).
Pursuant to the ETSI IPR Policy, no investigation, including IPR searches, has been carried out by ETSI. No guarantee
can be given as to the existence of other IPRs not referenced in ETSI SR 000 314 (or the updates on the ETSI Web
server) which are, or may be, or may become, essential to the present document.
Foreword
This Technical Report (TR) has been produced by ETSI Technical Committee Environmental Engineering (EE).
Introduction
The need to reduce emissions and to include energy efficiency as a new paradigm of industrial development is widely
acknowledged. In this context the mobile industry is making efforts to deploy energy efficient networks. In the Mobile
Green Manifesto 2012 it is stated that mobile industry will reduce its Green House Gas (GHG) emissions per every
connection by 40 % by 2020. Even if in this forecast all the variables in the mobile context are taken into account
altogether, it is estimated that 80 % of the energy consumption and GHG emissions of the mobile scenario are due to
networks. Studies and recommendations in this context are then highly valuable.
Moreover, it is reported from other sources (see [i.1]) that 3 % of current world-wide energy consumption is due to ICT,
which causes about 2 % of the overall CO2 emissions. Since mobile broadband data usage has experienced significant
growth and a thousand time expansion is expected by 2020, the consequence could be a significant increase of power
consumption of mobile networks.
The present document addresses mobile network level Radio Access energy efficiency measurements, and aims to
assess the complexity of these measurements both in real networks and in laboratory environments. The focus is on
mobile radio access networks, even if the alignment with Network Efficiency measurements in other contexts is deemed
as appropriate. Moreover the analysis is based on "partial" networks (as they are defined in the present document), with
an hint on how to extend the results to wider ("global") networks.
As for measurements in real networks, the report will study the most appropriate models to describe energy issues in
radio access networks and will introduce measurement definitions to validate these models. Data availability for models
and measurements will be checked.
As for measurements in laboratory environments, the report will consider the level of complexity that energy efficiency
measurements will impose, under the assumption they are an extension of pre-existing single node energy efficiency
measurements (like those defined in [i.2]).
The need for network level energy efficiency measurements is widely acknowledged, both as an extension of energy
efficiency evaluations on single nodes and as a "building block" for energy efficiency estimations of entire
communications networks.
On one hand, there are radio access features whose impact on energy efficiency cannot be fully estimated while
considering single node measurements only. As examples, we can cite RRM procedures, interference management,
Coordinated MultiPoint (CoMP), relay nodes management, heterogeneous network deployment, DTX methods, and
D2D/M2M techniques.
On the other hand, the radio access network is considered to be one of the most significant contributors to the energy
consumption of the entire communications network due to its extensive deployment. To form a complete view of
network energy efficiency, the radio access portion has to be considered properly and correct estimation of its energy
consumption is needed.
ETSI
6 ETSI TR 103 117 V1.1.1 (2012-11)
The present document, in defining and describing network level energy efficiency, can be beneficial for different
reasons:
• get an accurate measurement (rather than a statistical estimate) of radio network energy consumption and
efficiency;
• help radio access network operators understand and consequently improve the energy efficiency of their
networks, taking into consideration also the quality of service and quality of experience from the served users;
• enable radio access equipment vendors to demonstrate and improve energy efficiency features under real
conditions.
Throughout the present document it has to be pointed out that none of the statements made here are intended to specify
models, metrics or measurements procedures, but just to report them from a theoretical point of view, aiming to
highlight the feasibility of these various issues. Every definition of these issues will be possibly performed when and if
a different phase will be started in the Group, aiming at a Technical Specification purpose. Consider that some topics
dealt with in the present document could need more detailed analysis in case of an evolution towards a Technical
Specification.
The present document is organized as follows. First of all an analysis of the possible metrics to be adopted for network
level energy efficiency (clause 5) is presented, including an overview of the data available to build these metrics under
real network conditions (power consumed and throughput information). Clause 6 presents the measurement methods in
real networks as far as the studies have identified them so far, either from network inspection or relying on the users
data. Clause 7 reports the outcome of the evaluations made during the Technical Report lifetime about laboratories
network level measurements, with the recommendations on which an agreement has been reached. Next, in clause 8, a
method to extend the measurements made in the framework of partial network scenarios, following in particular the
ideas developed within the FP7 European Project EARTH and the method called E3F elaborated therein, is presented,
with an example about how to have figures of energy efficiency estimation in global networks. Finally a clause about
Recommendations for future work is given and the Report is completed by annexes dedicated to the main European
projects helpful in the preparation phase of the work (namely, EARTH and OPERA-Net).
ETSI
7 ETSI TR 103 117 V1.1.1 (2012-11)
1 Scope
The present document is about network level energy efficiency measurements, aiming to assess the complexity of these
measurements in real networks and in laboratory environments for the radio access networks framework. The radio
access is deemed to be the most relevant energy consumption source in mobile networks. In particular the focus of the
present document is on the "partial" radio access networks, representing a part of the whole network where the tests are
to be performed; possible extension to wider networks ("global" in the following) is taken into consideration as well. In
the networks analysed in the present document the considered elements are the so called "radio base stations", including
any possible representation of them, but excluding radio network controllers, backhauling or any equipment connecting
nodes to the core network. Energy efficiency of the User Equipment is not considered as well. Availability of data
useful to estimate energy efficiency in the network is dealt with as well in the present document. The content of the
present document is not intended to specify models, metrics or measurements procedures, rather to report them from an
analytical point of view, aiming to highlight the feasibility and effectiveness of the various options.
The radio access technologies considered in the report under study are mobile networks such as GSM, WCDMA and
LTE.
The present document also includes recommendations on possible future standardization work in the domain of wireless
network energy efficiency assessment.
2 References
References are either specific (identified by date of publication and/or edition number or version number) or
non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the
reference document (including any amendments) applies.
Referenced documents which are not found to be publicly available in the expected location might be found at
http://docbox.etsi.org/Reference.
NOTE: While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee
their long term validity.
2.1 Normative references
The following referenced documents are necessary for the application of the present document.
Not applicable.
2.2 Informative references
The following referenced documents are not necessary for the application of the present document but they assist the
user with regard to a particular subject area.
[i.1] KTHGenoa: "Energy Efficiency Measurements at the Network Level", Anders Västberg, Mats
Nilsson, Guowang Mio, Claes Beckman, Ki Won Sung, Jens Zander, 1st ETSI TC EE Workshop,
20-21 June 2012, Genoa Italy.
[i.2] ETSI TS 102 706: "Environmental Engineering (EE) Measurement Method for Energy Efficiency
of Wireless Access Network Equipment".
[i.3] "Energy efficiency KPIs for network measurement", Erik Friman, Ericsson; 1st ETSI TC EE
Workshop, 20-21 June 2012, Genoa Italy.
[i.4] OPERA-Net project information.
NOTE: Available at http://www.celtic-initiative.org/Projects/Celtic-projects/Call5/OPERA-Net/operanet-
default.asp.
ETSI
8 ETSI TR 103 117 V1.1.1 (2012-11)
[i.5] OPERA-Net2 project information.
NOTE: Available at http://www.celtic-initiative.org/Projects/Celtic-Plus-Projects/2011/OPERA-NET2/operanet2-
default.asp.
[i.6] L.Saker, S-E Alayoubi (FT) & H-O Scheck (NSN): "System selection and sleep mode for energy
saving in cooperative 2G/3G networks", VTC fall 2009, Alaska, September 09.
[i.7] Gilbert Micallef, et. al (NSN): "Energy efficiency evolution of mobile networks", VTC spring
2011.
[i.8] EARTH project summary leaflet.
NOTE: Available at www.ict-earth.eu.
[i.9] Project Summary.
NOTE: Available at https://www.ict-earth.eu/news/events/2012-07-06-
EARTH%20Project%20Presentation%20SummaryFNMS_12.pdf
[i.10] EARTH public deliverables.
NOTE: Available at https://www.ict-earth.eu/publications/deliverables/deliverables.html.
[i.11] "How Much Energy is Needed to Run a Wireless Network?", G. Auer et al. IEEE Wireless
Communications Magazine Special Issue on Technologies for Green Radio Communication
Networks. IEEE, Oct 2011.
[i.12] "How much energy is needed to run a wireless network?" G. Auer et al., in "Part I:
Communications Architectures and Models for Green Radio Networks" of "Green Radio
Communication Networks". Cambridge University Press, to appear Mai 2012.
[i.13] "Flexible power modeling of LTE base stations", C. Desset, B. Debaillie, V. Giannini, A. Fehske,
G. Auer, H. Holtkamp, W. Wajda, D. Sabella, F. Richter, M.J. Gonzalez, I. Gódor, (P.
Skillermark), M. Olsson, M. Imran, A. Ambrosy, O. Blume, IEEE WCNC 2012, Paris, April 2012.
[i.14] "Base Station Power Model", NTT DOCOMO, Alcatel-Lucent, Alcatel-Lucent Shanghai Bell,
Ericsson, Telecom Italia, R1-114336, TSG-RAN WG1 #67 San Francisco, USA,
November 14-18, 2011.
[i.15] "How green is my network? The E3F Energy Efficiency Evaluation Framework". O. Blume,
EARTH Training Workshop, Grenoble, June 13th, 2012.
NOTE: Available at
https://www.ict-earth.eu/news/events/2012-06-EARTH%20Training_WS_Estimation_Savings.pdf
[i.16] GSMA benchmarking service.
NOTE: Available at http://www.gsma.com/publicpolicy/mobile-energy-efficiency/mobile-energy-efficiency-
resources/.
[i.17] 3GPP TR 36.814: "Evolved Universal Terrestrial Radio Access (E-UTRA); Further advancements
for E-UTRA physical layer aspects".
[i.18] ITU-T Recommendation P.1201: "Parametric non-intrusive assessment of audiovisual media
streaming quality".
[i.19] ITU-R Report M.2135: "Guidelines for evaluation of radio interface technologies for
IMT-Advanced".
[i.20] 3GPP TR 36.805: "3rd Generation Partnership Project; Technical Specification Group Radio
Access Network; Study on Minimization of drive-tests in Next Generation Networks".
[i.21] 3GPP TSG-RAN WG2 Meeting #78: "Summary of [77bis#21] Joint/MDT: Scheduled IP
Throughput measurement scope".
ETSI
9 ETSI TR 103 117 V1.1.1 (2012-11)
[i.22] 3GPP TS 36.314 change request CR 0027: "Introduction of MDT measurements", TSG-RAN
WG2 Meeting #79.
NOTE: Available at http://www.3gpp.org/ftp/tsg_ran/WG2_RL2/TSGR2_79/Docs/R2-124358.zip.
[i.23] 3GPP TS 37.320 change request CR 0046: "Updates for MDT enhancements", TSG-RAN2
Meeting #79.
NOTE: Available at http://www.3gpp.org/ftp/tsg_ran/WG2_RL2/TSGR2_79/Docs/R2-124353.zip.
3 Definitions and abbreviations
3.1 Definitions
For the purposes of the present document, the following terms and definitions apply:
activity level: reflection of the amount of time (as a percentage of the total) during which data is generated by the
server and sent to the UE (or UE group) with the highest path loss
busy hour: period within a given 24-hour period during which the maximum RBS total load in a given 24-hour period
occurs
busy hour load: average RBS load during the busy hour
distributed RBS: RBS architecture which contains radio heads (RRH) close to antenna element and a central element
connecting RBS to network infrastructure
energy efficiency: relation between the useful output and energy/power consumption
Integrated RBS: RBS architecture in which all RBS element are located close to each other for example in one or two
cabinets
NOTE: The integrated RBS architecture may include Tower Mount Amplifier (TMA) close to antenna.
low load: average RBS load during time when there is very low traffic in network
medium term load: defined RBS load level between busy hour and low load levels
power consumption: power consumed by a device to achieve an intended application performance
power saving feature: feature which contributes to decreasing power consumption compared to the case when the
feature is not implemented
Radio Base Station (RBS): network component which serves one cell or more cells and interfaces the user terminal
(through air interface) and a wireless access network infrastructure
telecommunication network: network operated under a license granted by a national telecommunications authority,
which provides telecommunications between Network Termination Points (NTPs)
wireless access network: telecommunications network in which the access to the network (connection between user
terminal and network) is implemented without the use of wires
3.2 Abbreviations
For the purposes of the present document, the following abbreviations apply:
AC Alternating Current
ACK See TCP acronym.
BCCH Broadcast Control CHannel
BH Busy Hour
BS Base Station
BSC Base Station Controller
ETSI
10 ETSI TR 103 117 V1.1.1 (2012-11)
BTS Base Transceiver Station
CCH Common Channel
CDF Cumulative Distribution Function
CELTIC Industry-driven European research initiative
CPICH Common PIlot Channel
CRAN Cloud RAN (or C-RAN)
CS Circuit Switched
DC Direct Current
DL DownLink
DPCH Dedicated Physical Channel
DPI Direct Packet Inspection
DSP Digital Signal Processing
DTX Discontinuous Transmission
EARTH Energy Aware Radio and neTwork tecHnologies
EDGE Enhanced Datarate GSM Evolution
EUREKA European Research Coordination Agency
eUTRAN Evolved UMTS Terrestrial Radio Access Network
FIN See TCP acronym.
GBR Guaranteed Bit Rate
GERAN GSM/EDGE Radio Access Network
GGSN Gateway GPRS Support Node
GHG GreenHouse Gas
GSM Global System for Mobile communication
GSMA GSM Association
HSDPA High Speed Downlink Packet Access
HSPA High Speed Packet Access
HTTP HyperText Transfer Protocol
HVGA Half-size Video Graphics Array
HW HardWare
ICT Information Communications Technology
IP Internet Protocol
ITU International Telecommunications Union
JCNC Joint Channel and Network Coding
KPI Key Performance Indicator
LTE Long Term Evolution
MDT Minimization of Drive Tests
MEE Mobile Energy Efficiency
MIMO Multiple Input Multiple Output
MOS Mean Opinion Square
NA Not Applicable
NMS Network Management System
O&M Operation & Maintenance
OAM Operation And Maintenance
OEM Original Equipment Manufacturer
OPERA-Net Optimizing the Power Efficiency of Radio Access Networks
OS Operative System
OSS Operations Support Systems
PA Power Amplifier
P base band power consumption
BB
PDF Probability Distribution Function
PER Peak Error Rate
PM Performance Management
P power consumption of the digital processing unit
P
P maximum RF power
RF
PS Packet Switched
P fixed power consumption of the radio module
TRX
QCI QoS Class Identifier
QCIF Quarter Common Intermediate Format
QoE Quality of Experience (end-user)
QoS Quality of Services
RAB Radio Access Bearer
ETSI
11 ETSI TR 103 117 V1.1.1 (2012-11)
RAN Radio Access Network
RAT Radio Access Technology
RBS Radio Base Station
RF Radio Frequency
RNC Radio Network Controller
RRH Remote Radio Head
RRM Radio Resource Management
RX Receiver
SGW Serving Gateway
SME Small and medium-sized enterprises
SOTA State Of The Art
SW SoftWare
SYN See TCP acronym
TCP Transmission Control Protocol
NOTE: ACK, SYN and FIN are signalling in the TCP session.
TRX Transceiver
TV TeleVision
TX Transmitter
UDP User Datagram Protocol
UE User Equipment
UL UpLink
UMTS Universal Mobile Telecommunications System
UTRAN UMTS Terrestrial Radio Access Network
WCDMA Wideband Code Division Multiple Access
WP Work Package
XML Extensible Markup Language
4 Definition of the network in the scope
The overall positioning of the present document is graphically represented in Figure 1.

Figure 1: Graphical representation of WI scope
ETSI
12 ETSI TR 103 117 V1.1.1 (2012-11)
Specifically, the present document will build upon existing ETSI Standards, such as TS 102 706 [i.2] for single node
energy efficiency, evolving towards the definition of energy efficiency models and metrics for small portions of the
radio access networks. The commonly accepted definition of "access network" is adopted in the present document,
consisting of that part of the mobile network that connects the User Equipment to the Core Network (GERAN in
GSM/EDGE, UTRAN in WCDMA, eUTRAN in LTE). For the sake of simplicity and to build upon TS 102 706 [i.2]
the BSCs in GSM/EDGE networks and the RNCs in WCDMA networks are not considered in the tests and
measurement methods for energy efficiency purposes, nor are any backhaul segments or contributions from User
Equipment.
A graphical representation of what is defined in the present document as the "access network" is shown in Figure 2,
where the RNC/BSC/SGW are drawn for completeness but are out of the scope of the energy efficiency evaluations
within the present document.
Figure 2: Graphical representation of the access network under study
Evolving from small parts of the access networks proper extrapolation can be attempted towards the definition of
Energy Efficiency for complete access networks, such as the network of an operator in a country or anyway an
aggregation in a wide scale framework.
5 Metrics and data for network level energy efficiency
In this clause the issue of which kind of metrics could be the most suited one for network energy efficiency
measurements, in the particular case of the network in the scope of the present document, is dealt with. A definition of
proper metrics that could be followed in the measurement tests is given in the clause.
5.1 Metrics and energy efficiency in literature
An important objective in the design of cellular systems is to maximize the number of bits that can be delivered over a
certain time and in a given bandwidth; in this sense the throughput delivered in a given area and in such a bandwidth is
deemed as an appropriate way to estimate the performance of a network. The operation of the network is not required to
maintain overall throughput but it is also required to keep control of the perceived user quality. However quantifying
the end user's perceived quality may sometimes be difficult since the suitable metric typically depends on the service or
application being accessed.
Energy is considered a scarce resource that should be carefully utilized and thus energy consumption is considered in
the analysis. But the spectrum efficiency and the quality aspects are still important, so the analysis is a multi-dimension
one and it will sometimes be needed to trade one desired property for another, e.g. quality versus energy consumption.
ETSI
13 ETSI TR 103 117 V1.1.1 (2012-11)
In 3GPP quality performance metrics have been defined (see [i.17]), both for the case of full buffer traffic and for the
case of variable traffic. In particular, for the former case, the metrics that are suggested in 3GPP are:
• mean user throughput;
• throughput CDF;
• median and 5th percentile (worst) user throughput.
For variable traffic instead the 3GPP suggests as possible metrics:
• mean, 5th, 50th, 95th percentile user throughput;
• served (cell) throughput;
• harmonic mean normalized cell throughput;
• normalized cell throughput;
• resource utilization.
In TS 102 706 [i.2] for RBSs the "quality" is taken into account by ensuring a sufficient level of throughput for cell
edge users in the context of the dynamic measurement set-up; if the cell edge users do not achieve a level of throughput
above a given threshold the measurement is not considered. The traffic model that has been considered in the dynamic
measurement set-up is not the full buffer model as it considers users that have alternated ON and OFF periods.
In available literature (see as a possible example, EARTH deliverable D2.2 [i.10]), three main metrics definitions can be
found, depending on the focus of the measurement (which target to be addressed) and on which environment (urban,
rural, etc.) the measurement is to be applied:
I
• successfully transferred data volume per unit time over consumed power (ɛ );
A
• area unit over consumed power (ɛ );
S
• number of users over consumed power (ɛ );
and their definition is generally as follows:
Data _Volume _ per _ sec
bps
⎡ ⎤
ε = in
I
⎢ W⎥
Power ⎣ ⎦
Coverage _ Area
m
ε = in []
A
W
Power
Number _ Of _Users _ per _ sec
n_users
⎡ ⎤
ε = in
S
W ⋅s
⎢ ⎥
Power ⎣ ⎦
I
Data volume per unit time over power consumed (ɛ ) takes into consideration the achieved information exchange versus
I
the consumed power in the area where ɛ is computed. It has to be taken into consideration that there is power
consumption even if traffic is null (note that the metric is always zero when there is no traffic). This metric is thus
relevant for high traffic situations and for capacity-limited settings (urban areas). Only traffic of satisfied users, namely
users with a level of QoS/QoE above pre-defined thresholds, is to be taken into account as the traffic of unsatisfied
users is useless (see note in clause 5.2).
A
Area unit over consumed power (ɛ ) is otherwise important at low loads where systems are coverage limited and it is
therefore useful in case rural area scenarios. If the coverage area is kept fixed this metric is very reliable in terms of
consumed power and comparisons are easy.
ETSI
14 ETSI TR 103 117 V1.1.1 (2012-11)
S
Number of users per unit time over consumed power (ɛ ) could be an easy metric to be computed since the number of
users is quite an easy number to determine even if it is variable over time; however this metric has to consider satisfied
users (i.e. with a given QoS). A drawback of this metric is that it does not take into account the services and
applications that are being accessed: voice, video and data users are given the same weight, and different sessions
belonging to the same user are duplicated (may be an issue with smartphones).
An alternative metric could be the following (subject to further investigation):
Number _ Of _ Simult _ Scheduled _Users
n_users
⎡ ⎤
ε = in
SS
W
⎢ ⎥
⎣ ⎦
Power
The use of instantaneous data samples may be another approach considering that sampled measurements can be
averaged without impact on the measurement units.
5.2 Proposal of metrics for network level energy efficiency
The above defined metrics can be used to estimate Energy Efficiency together with the total energy consumption of the
I S
network. Focusing in particular to ɛ and ɛ , as defined in previous clause the first metric addressing the energy
efficiency as a ratio between offered throughput and consumed power in the area, the second as a ratio between number
of users and consumed power to serve them, they can be written, in a network area and highlighting the effect of
satisfied and unsatisfied users.
w
∑ i
i∈U
QoS
ε =
I
P
∑ BS ,i
i∈d
BS
card(U )
QoS
ε =
S
T × P
∑
BS ,i
i∈d
BS
i QoS
where w is the throughput per user with a given minimum QoS, U is the set of users with a given minimum QoS in
QoS
the area of the measurements represented by a number d of nodes, U' is the set of users that do not have the
BS
minimum QoS requirement in the area of the measurements; card(U) is defined as the cardinality of set U. T is the
observation time.
The proposed network metrics reflecting ε defined in clause 5.1 would be:
SS
card(U )
QoS
ε =
SS
P
∑
BS ,i
i∈d
BS
With regards to QoS, recognizing that it is difficult to specify a common QoS target for all kinds of users, QoS targets
will have to be set class by class:
1) For all kinds of classes, users are not satisfied if they are blocked or dropped. For circuit switched services like
voice or constant bit rate data, blocking and dropping are the only QoS measures.
2) For data services in HSDPA and LTE, an additional QoS metric is the throughput. We can differentiate
between GBR services (like streaming) and non-GBR ones (like web browsing users). If a GBR is specified
for the user, it is counted as satisfied by its throughput if it attains its GBR. Users that do not have a specified
GBR are considered as always satisfied by their throughput as these are best effort users (setting a target
throughput for them is too random to be considered in a metric).
ETSI
15 ETSI TR 103 117 V1.1.1 (2012-11)
The definitions of ε and ε (stated in the formulas above) are based on the introduction of QoS for all users, to
I S
ensure the fairness of such measurements. The adoption of one of the above mentioned metrics will in general depend
on a set of purposes, e.g. the aim of the test, the environment in which it is executed, or the availability of data. The
suggested adoption of ε and ε is based on the objective to include QoS information in the metric, and depends very
I S
heavily on the availability of information to build such a metric. In the case in which this information is not available
the differentiation among satisfied and unsatisfied users will not be possible and the traditional definition of metrics
I S
(namely ɛ or ɛ ) could be used, even if this will be a subject of further discussion and research.
NOTE: Increasing base station load levels,ε or ε may remain constant although the number of unsatisfied
S SS
users is increasing. When this metric is used to determine a base station's capacity limit, a possible
solution could then be to explicitly introduce throughput for unsatisfied users as a negative energy
efficiency element as defined below:
card(U ) +η card(U ' )
QoS QoS
ε =
S
T × P
∑ BS ,i
i∈d
BS
card(U ) +η card(U ' )
QoS QoS
ε =
SS
P
∑ BS ,i
i∈d
BS
QoS
where η is in the range [-1,0] and U' is the set of users in the measurement area for who the minimum
QoS requirement is not satisfied.
This negative impact is already implicitly taken into account in ε as the throughput provided to
I
unsatisfied users is taken from the satisfied users. For ε , energy efficiency will therefore start to
I
decrease with increasing number unsatisfied users.
6 Network level measurements in real networks
In clause 4 a definition of the network was provided; a network is a set of radio nodes in the Radio Access networks of
mobile Radio Access Technologies, such as GSM, WCDMA and LTE. The objective of this clause is to describe how
to perform energy efficiency measurements in such networks, according to the metrics and the constraints introduced in
clause 5.
In particular this clause is dedicated to real network measurements (specifically measurements performed in operating
networks loaded with real commercial traffic).
The measurement procedure to use will depend on the network information which can be collected and accessed for
further analysis. One possible option is based on the nodes knowledge of user data throughput and making that
information available to the operators. Another option is based on live test-users that perform real measurements in the
operators' networks to estimate the efficiency by means of the parameters defined in the energy efficiency metrics.
The potential measurement methods require a selection of the network area in which to perform the measurements
(according to the scope identified in clause 4), a selection of the quality of service threshold to ensure only valid users
are considered (according to the metric definitions in clause 5), and a step-by-step procedure to be followed to perform
the measurement.
6.1 Measurement procedure
Even if further analysis is needed in future stages, a procedure to make an energy efficiency measurement of real
networks is introduced in this clause, with further details provided in clauses 6.2 to 6.4.
ETSI
16 ETSI TR 103 117 V1.1.1 (2012-11)
In particular, for real network measurements, the following steps are foreseen:
• definition of the area of the measurement. As stated the metric to evaluate energy efficiency should be built in
a "partial" network area, according to the definition of the scope reported in clause 4. The definition of this
area is the first issue to be taken into consideration. According to the extension to global area, as it will be
described in clause 8, this partial area will be chosen in any of the geographical/sociological scenarios where
the metric could be used;
• observation time. According to the metrics definition (see clause 5) a proper observation time should be
introduced in order to capture the energy efficiency parameters of interest. This time should be long enough to
include any possible variations that could affect throughput and power consumption (such as build-in energy
efficiency measures in equipment), or changes in the number of users. In case of extension to global scale the
measurement should be repeated, with the same observation time, in different periods of day and in different
days in the year according to the load levels of the live networks to be investigated.
• power consumption measurement in the area. Within the framework of the above defined area and of the
defined observation time the overall consumed power has to be measured, collecting information from all the
equipment included in the mentioned area. For more details on this refer to clause 6.4.
• QoE throughput and/or "satisfied" users estimation. Always making reference to the area of the chosen
"partial" network and within the observation time constraints it has to be computed the overall throughput
and/or the number of users that could be defined as "satisfied" or "unsatisfied" according to the discussions
kept in clause 5. For this estimation many different options are available, ranging from inspection of the
network to the case of test users chosen to measure the network performance, and these options are described
in clause 6.3.5.
6.2 Definition of the area of the measurement
In this clause a possible definition of the area to be considered as the reference for the measurement is attempted. A
clause for reference is included of the area definition used in simulations. The partial area definition proposal is outlined
but further definition needs to be agreed upon in further specification work. Additional details are therefore difficult to
specify.
6.2.1 Area of measurement used in simulations
In simulations the concept of hexagonal tiers used to define reference areas; for measurement purposes the same
concept could apply. Two "tiers" to be introduced:
• "small" tier made by 7 sites (the central one and 6 around – red in the figure);
• "large" tier made by 19 sites (the central one and 18 around – blue and red in the figure).

Figure 3: Standard homogenous hexagonal layout used for network simulations
It is worth noting that the above mentioned area definition is a very theoretical
...